{"id":"W4411442784","doi":"10.1016/j.nlp.2025.100163","title":"Can “consciousness” be observed from large language model (LLM) internal states? Dissecting LLM representations obtained from Theory of Mind test with Integrated Information Theory and Span Representation analysis","year":2025,"lang":"en","type":"article","venue":"Natural Language Processing Journal","topic":"Language and cultural evolution","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"World Anti-Doping Agency","funders":"","keywords":"Representation (politics); Consciousness; Test (biology); Span (engineering); Cognitive science; Psychology; Integrated information theory; Cognitive psychology; Computer science; Engineering; Structural engineering; Political science; Geology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002015521,0.0002648101,0.0002862552,0.001461962,0.0002868921,0.001952612,0.0004087049,0.0004094086,0.0021881],"category_scores_gemma":[0.03307423,0.0002345083,0.0004174799,0.0008660471,0.001537823,0.003166428,0.001805573,0.0008788003,0.0001971313],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004814854,"about_ca_system_score_gemma":0.0004540947,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007450769,"about_ca_topic_score_gemma":0.0006332902,"domain_scores_codex":[0.9989912,0.0003689204,0.00008256698,0.0002429977,0.0002266984,0.00008766732],"domain_scores_gemma":[0.9881482,0.007115724,0.001757737,0.001944353,0.0006597149,0.0003743963],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001335716,0.0003177797,0.363525,0.0005829902,0.0005993435,0.0004621168,0.0170176,0.01794564,0.1794099,0.0934617,0.001120355,0.3242219],"study_design_scores_gemma":[0.00004118028,0.0008617002,0.6067768,0.000101488,0.000247303,0.0007791055,0.004820575,0.1360276,0.03813196,0.209692,0.002292841,0.0002274251],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9412333,0.000065991,0.05486851,0.0001482846,0.0000119811,0.00002566184,0.0001725927,0.0001434337,0.003330193],"genre_scores_gemma":[0.9923885,0.00001856215,0.00727704,0.00002056654,0.000003294948,0.0000242955,0.0001359703,0.00002601069,0.0001056213],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.0021881,"threshold_uncertainty_score":0.01065922,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01070362742570176,"score_gpt":0.3059250177171058,"score_spread":0.2952213902914041,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}