{"as_of":"2026-08-15T10:35:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:3ae6dfaae633e0dd0aca2a471a2f8ffcabf194c0a963fb5b47af0973af2ad150","coverage":[{"denominator":47,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":47,"source":"paper_references, paper_reference_links","source_observed_at":"2026-07-15T10:56:54.155632Z","state":"measured"},{"denominator":49,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":49,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-15T06:32:42.880941+00:00","state":"measured"},{"denominator":2,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":2,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-02T07:54:47.701071Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2606.04115","snapshot_observed_at":"2026-07-12T06:20:07.112455Z","title":"dMX: Differentiable mixed-precision assignment for low-precision floating-point formats.arXiv preprint arXiv:2606.04115, 2026","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.02893","last_updated":"2026-07-03T02:42:04Z","snapshot_observed_at":"2026-08-13T03:34:37.773819Z","submitted_at":"2026-07-03T02:42:04Z","title":"Variable Bit-width Quantization: Learning Per-Group Precision for \"Bigger-but-Smaller\" Language Models","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-07-12T06:20:07.112455Z"},"links":{"cited_paper":"/paper/2606.04115","citing_paper":"/paper/2607.02893"},"observation_digest":"sha256:ce037f1d068f4622e505be677c1f081c6820401d6cc06941c376533eaef9407e","observation_id":"f96a24be-7752-4512-886e-858bc8952ebf","resolution":{"observed_at":"2026-07-12T06:20:07.112455Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2606.04115","snapshot_observed_at":"2026-08-02T07:54:47.701071Z","title":"arXiv preprint arXiv:2606.04115 , year =","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.09800","last_updated":"2026-07-28T02:32:08Z","snapshot_observed_at":"2026-08-15T00:29:32.314814Z","submitted_at":"2026-07-09T15:23:42Z","title":"Reference Traces for Auditing Invisible Weight Updates and Guiding Exact-Budget Protection","version":3},"reference_index":76,"source":"arxiv_source","source_observed_at":"2026-08-02T07:54:47.701071Z"},"links":{"cited_paper":"/paper/2606.04115","citing_paper":"/paper/2607.09800"},"observation_digest":"sha256:1bfa189526d601fba91bf12916e7059bc400d9fc93b0e4eef3b9b3332e9954e3","observation_id":"28b9e92a-8fe5-4357-9de0-46d12ef0d38e","resolution":{"observed_at":"2026-08-02T07:54:47.701071Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"links":{"evidence":"/evidence","html":"/paper/2606.04115/citation-record","integrity":"/paper/2606.04115/integrity","json":"/paper/2606.04115/citation-record.json","paper":"/paper/2606.04115"},"outbound":[{"citation":{"cited_paper":{"arxiv_id":"2210.17323","last_updated":"2023-03-22T13:10:47Z","snapshot_observed_at":"2026-08-11T21:19:33.078277Z","submitted_at":"2022-10-31T13:42:40Z","title":"GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformers","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2210.17323","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"GPTQ: Accurate post-training quantization for generative pre-trained transformers.arXiv preprint arXiv:2210.17323, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2210.17323","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:34d422174f1b4b3efea1ddafed4cf3ddb0a4f83356a9c0591b861dc76afa2f76","observation_id":"ad2f2e7d-cf85-4525-b6c6-90bddde7b417","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"AWQ: Activation-aware weight quantization for on-device LLM compression and acceleration","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:23cab2da7695a7c959ebf1dad139d0d9a6bf53d0ad3066fbb3df44731ce3e452","observation_id":"ee30ae9b-ca06-4ee9-b672-3f21a2953f96","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"SmoothQuant: Accurate and efficient post-training quantization for large language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:095173864c6029a25c093541ec51a70e3d7c6bec2dff6064ad7daa205f1b6f99","observation_id":"01a4e7a5-eb51-4815-adf9-1e6d577bd502","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2303.17951","last_updated":"2023-06-15T08:14:02Z","snapshot_observed_at":"2026-08-13T12:12:37.645920Z","submitted_at":"2023-03-31T10:29:17Z","title":"FP8 versus INT8 for efficient deep learning inference","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.17951","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"FP8 versus INT8 for efficient deep learning inference.arXiv preprint arXiv:2303.17951, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2303.17951","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:4f9df2cf8dd1620be07fd71eac2abfbbccb7aae52dadf1db13d70b01686538d5","observation_id":"9278231b-8ec9-4fc8-a0cb-8329141765f4","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.10537","last_updated":"2023-10-19T16:38:33Z","snapshot_observed_at":"2026-08-13T05:48:18.482494Z","submitted_at":"2023-10-16T16:07:41Z","title":"Microscaling Data Formats for Deep Learning","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.10537","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Microscaling data formats for deep learning.arXiv preprint arXiv:2310.10537, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2310.10537","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:d420e0c373de0dead2c09503f9557bd8990b3807510aa7f2d844ddbc7b188a1c","observation_id":"481d94f3-5b30-4a98-b9f7-de44a7593944","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"OCP microscaling formats (MX) specification","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:6de0d3b01daac51133302fce2462e11c846886d3b7b3bc5c9ec214b4c4a221ee","observation_id":"93c15a61-5ed5-4e82-81bf-e6fad0f6021c","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Castro, Denis Kuznedelev, Andrei Panferov, Eldar Kurtic, Shubhra Pandit, Alexandre Marques, Mark Kurtz, Saleh Ashkboos, Torsten Hoefler, and Dan Alistarh","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:9fe73bd039af2eaeb50a5a4f8416d79ef2b6927455a77e3d89966424d7b93a47","observation_id":"db86112b-4290-4684-899a-0786cffaca66","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2601.22347","last_updated":"2026-05-27T21:46:34Z","snapshot_observed_at":"2026-08-03T06:35:09.390088Z","submitted_at":"2026-01-29T21:46:26Z","title":"Pushing the Limits of Block Rotations in Post-Training Quantization","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2601.22347","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"MixQuant: Pushing the limits of block rotations in post-training quantization","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2601.22347","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:e19a9e2b6ed598f60ac711b1e7e492d5793032800b012f115d89ca7fa8b42398","observation_id":"1c0758d5-1a34-4f28-a2f0-20785ac3b287","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Gradient-free training of quantized neural networks.arXiv preprint arXiv:2410.09734, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:217bfd83ae8ba3fc05e6d87ddd7a84dff36167b0a13cd08cd5782638fe4c678d","observation_id":"ab19d72d-a88c-444d-ac8a-d5061e50152c","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1812.00090","last_updated":"2018-11-30T23:15:45Z","snapshot_observed_at":"2026-08-14T17:50:18.985454Z","submitted_at":"2018-11-30T23:15:45Z","title":"Mixed Precision Quantization of ConvNets via Differentiable Neural Architecture Search","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1812.00090","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Mixed precision quantization of ConvNets via differentiable neural architecture search.arXiv preprint arXiv:1812.00090, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/1812.00090","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:8dee12427630edcd4895e9f9d36cd28a1112b78385f8593f8929654d7bddfcdf","observation_id":"c518c453-cef8-44b6-81de-20f6f8beec97","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"InfoQ: Mixed-precision quantization via global information flow","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:d93dd678a2ff5b038882e5ef1c0415207ce6fcc3075305f54071f8b2907d29cb","observation_id":"6549b819-c731-416b-bf69-6dd520ea3759","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Mix-QSAM: Mixed-precision quantization of the segment anything model","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:aa646f66ac24e4df7fc3567508fa7b3393c1572a8e506f09e7ec23bbc7899275","observation_id":"bf4d76be-c080-4b2d-9236-d0604f2a54b7","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Mahoney, and Kurt Keutzer","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:67c09f52b3d82486c02a96de87331e3ca958d359c9deaf0ab4407d874c04b733","observation_id":"08fc48cc-320a-482d-8498-3310774cc854","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Mahoney, and Kurt Keutzer","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:ed01f2376f213a3f0d7a25106aee1d3ca89c21193e988365dbfc3aca39f7704b","observation_id":"1643021f-e6ed-4749-98ba-6c151a89588f","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"FracBits: Mixed precision quantization via fractional bit-widths","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:0dd37ed5e0c4bfb2f2fa5e6eb68bc3cf6ac8d1714642469c325402413574cc8e","observation_id":"7c646271-eefb-491d-8b1b-b9ce184dc32b","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"SDQ: Stochastic differentiable quantization with mixed precision","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:863c4e30f8991e6e6a5b17d7c32567c173cea7efc5dd00abf8f33a008f831d83","observation_id":"6fe206fe-0942-48be-8593-3ff7846ecb38","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"BSQ: Exploring bit-level sparsity for mixed-precision neural network quantization","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:4cdcbd5c4112192efb040572ae84618b8bfe5c2a0dd9a706e94f70b3018fce43","observation_id":"19daa9d3-7fdc-4aa8-b93f-a7309fd79afa","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1308.3432","last_updated":"2013-08-15T15:19:34Z","snapshot_observed_at":"2026-08-14T04:51:04.817737Z","submitted_at":"2013-08-15T15:19:34Z","title":"Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1308.3432","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Estimating or propagating gradients through stochastic neurons for conditional computation.arXiv preprint arXiv:1308.3432, 2013","venue":null,"work_id":null,"year":2013},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/1308.3432","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:9a3c0083fa0e53ef075da0ff4dc489bdf0e545e155f44d07d0f4142f297e6b8a","observation_id":"cc6f69d7-5571-4f37-81e3-abd265593efe","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2405.16406","last_updated":"2025-02-20T06:07:00Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-05-26T02:15:49Z","title":"SpinQuant: LLM quantization with learned rotations","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2405.16406","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Spinquant: Llm quantization with learned rotations.arXiv preprint arXiv:2405.16406, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2405.16406","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:817247e7fb96c186bbeda45e8f2b861a417e33cb67d619671fb9adc55698bb17","observation_id":"a434e32c-6ecf-4863-8ea6-b9cfc6d9fd30","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2209.05433","last_updated":"2022-09-29T20:47:07Z","snapshot_observed_at":"2026-08-14T05:39:52.416140Z","submitted_at":"2022-09-12T17:39:55Z","title":"FP8 Formats for Deep Learning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2209.05433","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"FP8 formats for deep learning.arXiv preprint arXiv:2209.05433, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2209.05433","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:1dd1b4b569a6bde78f55217e85f721dc87736df6b0e80bd77b183529b903ddd5","observation_id":"d7d612a4-e4ea-4e67-87e4-9cb38f897235","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.21783","last_updated":"2024-11-23T23:27:33Z","snapshot_observed_at":"2026-08-13T17:20:44.002518Z","submitted_at":"2024-07-31T17:54:27Z","title":"The Llama 3 Herd of Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.21783","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"The Llama 3 herd of models.arXiv preprint arXiv:2407.21783, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2407.21783","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:2c2987ced0ef5c1eb30ebf8aa7202b5de7c18a9e670a7fb6948b149107e9b580","observation_id":"9d22b41c-d9de-467f-bbf7-03c1fdeb3fa1","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.09388","last_updated":"2025-05-14T13:41:34Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2025-05-14T13:41:34Z","title":"Qwen3 Technical Report","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.09388","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Qwen3 technical report.arXiv preprint arXiv:2505.09388, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2505.09388","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:71a27343b853614d16998ac39565f5de92766403cece9ae4b8d73d7bd8c83989","observation_id":"2d9dfe9b-da9c-4878-97f4-990864d39101","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2502.02737","last_updated":"2025-02-04T21:43:16Z","snapshot_observed_at":"2026-08-14T05:03:05.661970Z","submitted_at":"2025-02-04T21:43:16Z","title":"SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2502.02737","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"SmolLM2: When smol goes big – data-centric training of a small language model.arXiv preprint arXiv:2502.02737, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2502.02737","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:f3c6acba0b42dddec792e19569251a23576fd08b81dab2b769b434eec793d141","observation_id":"23221f78-ef63-498a-896a-112aaeb2f0c5","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.17557","last_updated":"2024-10-31T11:37:49Z","snapshot_observed_at":"2026-08-14T14:43:02.130172Z","submitted_at":"2024-06-25T13:50:56Z","title":"The FineWeb Datasets: Decanting the Web for the Finest Text Data at Scale","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.17557","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"The FineWeb datasets: Decanting the web for the finest text data at scale.arXiv preprint arXiv:2406.17557, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2406.17557","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:62826df05dda9f623cd1ec1aaca8b16f0ccac5b3013cecccf6dd4900378b2f4a","observation_id":"39fc2f7e-39a6-4012-8bd0-83889c053788","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Pointer sentinel mixture models","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:52e877fe35e217f3026f5f9021eb893e4e361dc635455cf3f0a32863ad0524c1","observation_id":"da55252b-2d7e-4545-b41f-7833a312284d","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1803.05457","last_updated":"2018-03-14T18:04:21Z","snapshot_observed_at":"2026-08-14T19:36:07.505691Z","submitted_at":"2018-03-14T18:04:21Z","title":"Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1803.05457","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Think you have solved question answering? try ARC, the AI2 reasoning challenge.arXiv preprint arXiv:1803.05457, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/1803.05457","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:8e4851c810c08df340261f03dc202bf89077b7d76f1acdfea3519753ec8c18a8","observation_id":"9d153449-e1ac-486a-92bb-8d07f182e904","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"HellaSwag: Can a machine really finish your sentence? InProceedings of the 57th Annual Meeting of the Association for Computational Linguistics, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:ef69e76c1080a145df9bd76acad7fdc0c830eb2ca9609d97fa1c2f8b6d2b8b66","observation_id":"0cd6ab72-3a53-4873-b04e-f88a3b0e613b","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"WinoGrande: An adversarial Winograd schema challenge at scale","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:ff9c38bec45d3c34e6a573f3154a5ac34ac71e3c1e3bcf0e556e3ad47e0fe1f6","observation_id":"003985f4-3b25-4b5d-80e7-de8a740a3f5b","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"LightEval: A lightweight framework for LLM evaluation, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:ad2abad8256bb75a2772ec3acd58a995e0a3da04b69177ab86094b3549bf2405","observation_id":"16bb2e4a-95a1-4fa8-9132-86d0a3aa8f71","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Xilinx/brevitas, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:9e33098122a142aa8522cc76cf202c91fde72a60c1a6886095f7fcb152456a36","observation_id":"d601c840-e235-4b08-99dd-ec3d9529dae1","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"HAQ: Hardware-aware automated quantization with mixed precision","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:e925697b5a4a08a8f0d59a70708ed24b21308b08d617f8b559bd45a64d012934","observation_id":"ed56cbed-bca2-42c5-bca4-78b413f7aead","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Mahoney, and Kurt Keutzer","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:faf3e9ec4ea5bbcaaf1c8e5a4f3060992e0818f4ced93ea23217f48ea619c83b","observation_id":"48f2a0df-601f-4909-b6f3-0a5c51f33691","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Towards mixed-precision quantization of neural networks via constrained optimization","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:193f43133affb8b0c318691d212e9e03ec86e617fb94f43f423d277451a356c8","observation_id":"7f4fbf9b-d0ac-4ec0-bd2f-40fafd3a5417","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"APTQ: Attention- aware post-training mixed-precision quantization for large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:835befc8102dfcd5d275fde23a8393270791c67aa89494c0df295c4992110f6d","observation_id":"4fac3b13-23c5-4ecd-b57b-956ee215fab9","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.14363","last_updated":"2025-02-03T21:45:32Z","snapshot_observed_at":"2026-08-13T00:10:37.377984Z","submitted_at":"2024-12-18T22:01:55Z","title":"ResQ: Mixed-Precision Quantization of Large Language Models with Low-Rank Residuals","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.14363","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"ResQ: Mixed-precision quan- tization of large language models with low-rank residuals.arXiv preprint arXiv:2412.14363, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2412.14363","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:ea078d6d435009f8c2900a52017474b472aff0ba21a68612179e4e53d23740fc","observation_id":"74e42f0f-80a9-4853-899d-dc176c6a364d","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Rethinking differentiable search for mixed-precision neural networks","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:8d25fdf86b05f775d30b2c3a7e238573087ab42d89569579279d7c4a6295e0ec","observation_id":"08cd2662-885c-4a08-aefa-9a18558cccb4","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2201.07703","last_updated":"2022-09-06T02:40:26Z","snapshot_observed_at":"2026-08-15T00:42:03.013477Z","submitted_at":"2022-01-19T16:43:17Z","title":"Q-ViT: Fully Differentiable Quantization for Vision Transformer","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2201.07703","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Q-ViT: Fully differentiable quantization for vision transformer.arXiv preprint arXiv:2201.07703, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/2201.07703","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:ce161b70a5ee21675c1a4f106990aa86a90a53b5e95e5bec5c496693d730d343","observation_id":"a7f2ba69-2e4f-4f6a-9051-69c02e0e1e90","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Jennings, and Arnon Netzer","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:28e99d9523161063fe3091228945ddf02f7e92deadfeb702b15a06b219867682","observation_id":"bde7d15b-b770-4eb4-86f7-c99623cd7404","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1611.01144","last_updated":"2017-08-05T22:45:19Z","snapshot_observed_at":"2026-08-15T03:05:41.956181Z","submitted_at":"2016-11-03T19:48:08Z","title":"Categorical Reparameterization with Gumbel-Softmax","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1611.01144","snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Categorical reparameterization with gumbel-softmax","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"cited_paper":"/paper/1611.01144","citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:6e6e2eed3a583f78aba4296417342d8045cadede0a6a7ed243785ff2bbef90b7","observation_id":"661e2b61-8681-47a3-b38e-7d6d9a005ae8","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Maddison, Andriy Mnih, and Yee Whye Teh","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:0d9f2618b5551b409656e4698e4b7310bd3f1bfe4c6831c9e4952e0d119ccc7b","observation_id":"7b0a9cef-c448-460e-8b09-dfe8ee63a201","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Bayesian bits: Unifying quantization and pruning","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:8971875f2f0ffa1446e7bad759e1fefcc844dfc8e099d010f0a025332db58ee4","observation_id":"ee03602b-45d2-4e47-be90-6742f83b1812","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Mixed precision DNNs: All you need is a good parametrization","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:a5c46fb2a6ff9419fb8ff1bfb0559e1543de3cb3003a2dabf0d92e6ed45e3897","observation_id":"071a89b9-b2e8-4bf8-9409-5df2ed95db63","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Micromix: Efficient mixed-precision quantization with microscaling formats for large language models","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:4afcdedbb9c822f018aac7ea855d8edacf23c447cb6ea80b6f51dd377369ca9e","observation_id":"9f154de3-d264-4d0b-85a0-0d9a09ac4ec0","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Mixture compressor for mixture-of-experts LLMs gains more","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:78b8f3d03f16916a70f7220c92498a7b6ba0711455b8184e24819ddebca1898f","observation_id":"1706284e-6580-47a0-8ce0-69f1fa70e68b","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Ieee standard for floating-point arithmetic.IEEE Std 754-2019 (Revision of IEEE 754-2008), pages 1–84, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:83d8c397d91d3b25a10f7bff666224de7dab250249f02dbb258afb7a89e3922e","observation_id":"8729cef7-133b-4cbb-8482-b21842caaf08","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"Jain, Albert Gural, Michael Wu, and Chris H","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:3f088d1a475005d413b73fc4d14809b29f2460bb0c2ec139da9ef541cf5e8960","observation_id":"94e0abd4-2429-4786-a96d-5360e50e7183","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-15T10:56:54.155632Z","title":"PyTorch: An imperative style, high-performance deep learning library","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats","version":2},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-07-15T10:56:54.155632Z"},"links":{"citing_paper":"/paper/2606.04115"},"observation_digest":"sha256:f192429a9154bb41f5a4d92a9a85e8022a7ba3ee686acb1e68fa6244a8e02e48","observation_id":"a0ec5a63-2236-4959-bb80-e4552ea9db58","resolution":{"observed_at":"2026-07-15T10:56:54.155632Z","resolver_source":null,"status":"malformed_identifier"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2606.04115","last_updated":"2026-07-14T14:04:32Z","latest_version":2,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-02T05:34:23.902031Z","submitted_at":"2026-06-02T18:23:20Z","title":"dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats"},"reference_resolution":{"displayed":47,"state_counts":{"malformed_identifier":1,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":46,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":47},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-15T06:32:42.880941+00:00","source":"crossref"},{"observed_at":"2026-08-15T06:32:39.529945+00:00","source":"retraction_watch"}],"thesis":"As of 15 August 2026, this Paper Citation Record lists 47 of 47 outbound references and 2 inbound Pith citation observations for arXiv:2606.04115."}