{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:G6LNC2HTXDNYVPF4N7DYYBG3PQ","short_pith_number":"pith:G6LNC2HT","schema_version":"1.0","canonical_sha256":"3796d168f3b8db8abcbc6fc78c04db7c3db02f9514eb496e3bec6fe156476151","source":{"kind":"arxiv","id":"2606.27199","version":1},"attestation_state":"computed","paper":{"title":"Forecasting With LLMs: Improved Generalization Through Feature Steering","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Bradford Levy, Humzah Merchant","submitted_at":"2026-06-25T15:59:20Z","abstract_excerpt":"Successful forecasting involves identifying patterns between historical and future states of the world which generalize to future observations. We apply LLMs to a variety of forecasting tasks and inspect their internal states using sparse autoencoders to understand whether they appear to rely on time-specific pieces of knowledge versus generalizable patterns. Our analyses identify features associated with both time-aware reasoning and look-ahead-biased reasoning. We then apply the LLMs to an entirely different domain and intervene on these features. We find that amplifying time-awareness featu"},"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":"2606.27199","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-25T15:59:20Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"2b890720ac3ff4085fa2477f7ab0915dc17a74d87aa6626bfba048b8e2e059e9","abstract_canon_sha256":"a8d778c9e7be0a2692dd23f52558abeaaf0f5cebd9470a7f53365bc8bb132162"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-26T01:16:13.415142Z","signature_b64":"BybG7uQRtj5AAwVLspcTA6LpiRmxIF91Pqv+j2m1fSAFvVC7VYR2VB2D3jAhxEPlgRY3390Na0qaY5A3TX9/Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3796d168f3b8db8abcbc6fc78c04db7c3db02f9514eb496e3bec6fe156476151","last_reissued_at":"2026-06-26T01:16:13.414764Z","signature_status":"signed_v1","first_computed_at":"2026-06-26T01:16:13.414764Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Forecasting With LLMs: Improved Generalization Through Feature Steering","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Bradford Levy, Humzah Merchant","submitted_at":"2026-06-25T15:59:20Z","abstract_excerpt":"Successful forecasting involves identifying patterns between historical and future states of the world which generalize to future observations. We apply LLMs to a variety of forecasting tasks and inspect their internal states using sparse autoencoders to understand whether they appear to rely on time-specific pieces of knowledge versus generalizable patterns. Our analyses identify features associated with both time-aware reasoning and look-ahead-biased reasoning. We then apply the LLMs to an entirely different domain and intervene on these features. We find that amplifying time-awareness featu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27199","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/2606.27199/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":"2606.27199","created_at":"2026-06-26T01:16:13.414820+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.27199v1","created_at":"2026-06-26T01:16:13.414820+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27199","created_at":"2026-06-26T01:16:13.414820+00:00"},{"alias_kind":"pith_short_12","alias_value":"G6LNC2HTXDNY","created_at":"2026-06-26T01:16:13.414820+00:00"},{"alias_kind":"pith_short_16","alias_value":"G6LNC2HTXDNYVPF4","created_at":"2026-06-26T01:16:13.414820+00:00"},{"alias_kind":"pith_short_8","alias_value":"G6LNC2HT","created_at":"2026-06-26T01:16:13.414820+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ","json":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ.json","graph_json":"https://pith.science/api/pith-number/G6LNC2HTXDNYVPF4N7DYYBG3PQ/graph.json","events_json":"https://pith.science/api/pith-number/G6LNC2HTXDNYVPF4N7DYYBG3PQ/events.json","paper":"https://pith.science/paper/G6LNC2HT"},"agent_actions":{"view_html":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ","download_json":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ.json","view_paper":"https://pith.science/paper/G6LNC2HT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.27199&json=true","fetch_graph":"https://pith.science/api/pith-number/G6LNC2HTXDNYVPF4N7DYYBG3PQ/graph.json","fetch_events":"https://pith.science/api/pith-number/G6LNC2HTXDNYVPF4N7DYYBG3PQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ/action/storage_attestation","attest_author":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ/action/author_attestation","sign_citation":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ/action/citation_signature","submit_replication":"https://pith.science/pith/G6LNC2HTXDNYVPF4N7DYYBG3PQ/action/replication_record"}},"created_at":"2026-06-26T01:16:13.414820+00:00","updated_at":"2026-06-26T01:16:13.414820+00:00"}