{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:5ZGAYXVIY7NXUG25FHVOGVVICA","short_pith_number":"pith:5ZGAYXVI","schema_version":"1.0","canonical_sha256":"ee4c0c5ea8c7db7a1b5d29eae356a8100f1bd6abdea42eb84efd5663e60ce735","source":{"kind":"arxiv","id":"2507.02092","version":1},"attestation_state":"computed","paper":{"title":"Energy-Based Transformers are Scalable Learners and Thinkers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV"],"primary_cat":"cs.LG","authors_text":"Alexi Gladstone, Aman Chadha, Ganesh Nanduru, Heng Ji, Hyeonjeong Ha, Jundong Li, Md Mofijul Islam, Peixuan Han, Tariq Iqbal, Yilun Du","submitted_at":"2025-07-02T19:17:29Z","abstract_excerpt":"Inference-time computation techniques, analogous to human System 2 Thinking, have recently become popular for improving model performances. However, most existing approaches suffer from several limitations: they are modality-specific (e.g., working only in text), problem-specific (e.g., verifiable domains like math and coding), or require additional supervision/training on top of unsupervised pretraining (e.g., verifiers or verifiable rewards). In this paper, we ask the question \"Is it possible to generalize these System 2 Thinking approaches, and develop models that learn to think solely from"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2507.02092","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-07-02T19:17:29Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CV"],"title_canon_sha256":"ebf17dc5ba129225c1ce95c9d232a61a6ea58856bc1235e1e4bfcf1bda33e823","abstract_canon_sha256":"628fd680bd1e5669031d031f61433635acd505a312b78a35b0423d54253445c3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:31:11.876735Z","signature_b64":"/g7H5Iwm/grjsmyjt5aYN0IYMlUIuAgbY5aznFaHei3mcF0TH+vR6IzXuXUBX8HaolayqkcdQLVc9/c7Co+NDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ee4c0c5ea8c7db7a1b5d29eae356a8100f1bd6abdea42eb84efd5663e60ce735","last_reissued_at":"2026-07-05T11:31:11.876250Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:31:11.876250Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Energy-Based Transformers are Scalable Learners and Thinkers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV"],"primary_cat":"cs.LG","authors_text":"Alexi Gladstone, Aman Chadha, Ganesh Nanduru, Heng Ji, Hyeonjeong Ha, Jundong Li, Md Mofijul Islam, Peixuan Han, Tariq Iqbal, Yilun Du","submitted_at":"2025-07-02T19:17:29Z","abstract_excerpt":"Inference-time computation techniques, analogous to human System 2 Thinking, have recently become popular for improving model performances. However, most existing approaches suffer from several limitations: they are modality-specific (e.g., working only in text), problem-specific (e.g., verifiable domains like math and coding), or require additional supervision/training on top of unsupervised pretraining (e.g., verifiers or verifiable rewards). In this paper, we ask the question \"Is it possible to generalize these System 2 Thinking approaches, and develop models that learn to think solely from"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.02092","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2507.02092/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2507.02092","created_at":"2026-07-05T11:31:11.876308+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.02092v1","created_at":"2026-07-05T11:31:11.876308+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.02092","created_at":"2026-07-05T11:31:11.876308+00:00"},{"alias_kind":"pith_short_12","alias_value":"5ZGAYXVIY7NX","created_at":"2026-07-05T11:31:11.876308+00:00"},{"alias_kind":"pith_short_16","alias_value":"5ZGAYXVIY7NXUG25","created_at":"2026-07-05T11:31:11.876308+00:00"},{"alias_kind":"pith_short_8","alias_value":"5ZGAYXVI","created_at":"2026-07-05T11:31:11.876308+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.18206","citing_title":"Fixed-Point Reasoners: Stable and Adaptive Deep Looped Transformers","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22726","citing_title":"Text Dictates, Music Decorates: Energy-based Attention for Editable Dance Motion Generation","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2602.01651","citing_title":"On the Spatiotemporal Dynamics of Generalization in Neural Networks","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03878","citing_title":"Learning 3D Reconstruction with Priors in Test Time","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2503.09567","citing_title":"Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Models","ref_index":212,"is_internal_anchor":false},{"citing_arxiv_id":"2412.06769","citing_title":"Training Large Language Models to Reason in a Continuous Latent Space","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10272","citing_title":"The Phase Is the Gradient: Equilibrium Propagation for Frequency Learning in Kuramoto Networks","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11403","citing_title":"One Scale at a Time: Scale-Autoregressive Modeling for Fluid Flow Distributions","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07588","citing_title":"Revisiting Transformer Layer Parameterization Through Causal Energy Minimization","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13521","citing_title":"C-voting: Confidence-Based Test-Time Voting without Explicit Energy Functions","ref_index":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA","json":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA.json","graph_json":"https://pith.science/api/pith-number/5ZGAYXVIY7NXUG25FHVOGVVICA/graph.json","events_json":"https://pith.science/api/pith-number/5ZGAYXVIY7NXUG25FHVOGVVICA/events.json","paper":"https://pith.science/paper/5ZGAYXVI"},"agent_actions":{"view_html":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA","download_json":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA.json","view_paper":"https://pith.science/paper/5ZGAYXVI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.02092&json=true","fetch_graph":"https://pith.science/api/pith-number/5ZGAYXVIY7NXUG25FHVOGVVICA/graph.json","fetch_events":"https://pith.science/api/pith-number/5ZGAYXVIY7NXUG25FHVOGVVICA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA/action/storage_attestation","attest_author":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA/action/author_attestation","sign_citation":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA/action/citation_signature","submit_replication":"https://pith.science/pith/5ZGAYXVIY7NXUG25FHVOGVVICA/action/replication_record"}},"created_at":"2026-07-05T11:31:11.876308+00:00","updated_at":"2026-07-05T11:31:11.876308+00:00"}