{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:OLK4YYG7XGGFJD2C4H6JNK54WM","short_pith_number":"pith:OLK4YYG7","schema_version":"1.0","canonical_sha256":"72d5cc60dfb98c548f42e1fc96abbcb3264309a9bd17cc24e2e8a5d583d09396","source":{"kind":"arxiv","id":"2605.26738","version":1},"attestation_state":"computed","paper":{"title":"KARMA: Karma-Aligned Reward Model Adaptation","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jared Scott, Jesse Roberts","submitted_at":"2026-05-26T09:12:14Z","abstract_excerpt":"Human communication depends on implicit social signals where effectiveness is shaped by tone, context, and conversational norms rather than semantic content alone. We introduce KARMA (Karma-Aligned Reward Model Adaptation), a framework for LLM learning of context-sensitive conversational behavior from large-scale social interaction data. KARMA trains a reward model on Reddit conversations to predict response valuation conditioned on context, and uses this signal to fine-tune language models via reinforcement learning to improve performance on pragmatics-mediated tasks. Critically, we find that"},"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":"2605.26738","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-26T09:12:14Z","cross_cats_sorted":[],"title_canon_sha256":"022ac5a9f5bafe3dd6ff5446999917b8251d0a264080dc0fa9a33c5da79bf04f","abstract_canon_sha256":"22c158c01a6562f9c5d393073e05527011edaeffe82a91a18ce031fc01e65f85"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:06:10.131118Z","signature_b64":"vkv0eut2N31UzB76+GqeIP0mpPE2goIgXRyjInz/6xrnFJytzEn7gVZiHz0HTurD9/WGCJOZw8IPdTX4PnfyDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"72d5cc60dfb98c548f42e1fc96abbcb3264309a9bd17cc24e2e8a5d583d09396","last_reissued_at":"2026-05-27T01:06:10.130548Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:06:10.130548Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"KARMA: Karma-Aligned Reward Model Adaptation","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jared Scott, Jesse Roberts","submitted_at":"2026-05-26T09:12:14Z","abstract_excerpt":"Human communication depends on implicit social signals where effectiveness is shaped by tone, context, and conversational norms rather than semantic content alone. We introduce KARMA (Karma-Aligned Reward Model Adaptation), a framework for LLM learning of context-sensitive conversational behavior from large-scale social interaction data. KARMA trains a reward model on Reddit conversations to predict response valuation conditioned on context, and uses this signal to fine-tune language models via reinforcement learning to improve performance on pragmatics-mediated tasks. Critically, we find that"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.26738","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/2605.26738/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":"2605.26738","created_at":"2026-05-27T01:06:10.130638+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.26738v1","created_at":"2026-05-27T01:06:10.130638+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.26738","created_at":"2026-05-27T01:06:10.130638+00:00"},{"alias_kind":"pith_short_12","alias_value":"OLK4YYG7XGGF","created_at":"2026-05-27T01:06:10.130638+00:00"},{"alias_kind":"pith_short_16","alias_value":"OLK4YYG7XGGFJD2C","created_at":"2026-05-27T01:06:10.130638+00:00"},{"alias_kind":"pith_short_8","alias_value":"OLK4YYG7","created_at":"2026-05-27T01:06:10.130638+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/OLK4YYG7XGGFJD2C4H6JNK54WM","json":"https://pith.science/pith/OLK4YYG7XGGFJD2C4H6JNK54WM.json","graph_json":"https://pith.science/api/pith-number/OLK4YYG7XGGFJD2C4H6JNK54WM/graph.json","events_json":"https://pith.science/api/pith-number/OLK4YYG7XGGFJD2C4H6JNK54WM/events.json","paper":"https://pith.science/paper/OLK4YYG7"},"agent_actions":{"view_html":"https://pith.science/pith/OLK4YYG7XGGFJD2C4H6JNK54WM","download_json":"https://pith.science/pith/OLK4YYG7XGGFJD2C4H6JNK54WM.json","view_paper":"https://pith.science/paper/OLK4YYG7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.26738&json=true","fetch_graph":"https://pith.science/api/pith-number/OLK4YYG7XGGFJD2C4H6JNK54WM/graph.json","fetch_events":"https://pith.science/api/pith-number/OLK4YYG7XGGFJD2C4H6JNK54WM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OLK4YYG7XGGFJD2C4H6JNK54WM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OLK4YYG7XGGFJD2C4H6JNK54WM/action/storage_attestation","attest_author":"https://pith.science/pith/OLK4YYG7XGGFJD2C4H6JNK54WM/action/author_attestation","sign_citation":"https://pith.science/pith/OLK4YYG7XGGFJD2C4H6JNK54WM/action/citation_signature","submit_replication":"https://pith.science/pith/OLK4YYG7XGGFJD2C4H6JNK54WM/action/replication_record"}},"created_at":"2026-05-27T01:06:10.130638+00:00","updated_at":"2026-05-27T01:06:10.130638+00:00"}