{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00101932,0.0001945372,0.0003217937,0.0003131687,0.0007472045,0.0005093968,0.0002693799,0.0001292284,0.0001520336],"category_scores_gemma":[0.002537581,0.000137171,0.0001120961,0.001154663,0.0001909235,0.001445173,0.00005974541,0.0005366907,6.625336e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001716588,"about_ca_system_score_gemma":0.0002801281,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.01026454,"about_ca_topic_score_gemma":0.01911264,"domain_scores_codex":[0.9979489,0.000502729,0.0005112005,0.0002561839,0.00047671,0.0003042278],"domain_scores_gemma":[0.9978194,0.000612911,0.0006091597,0.0001669362,0.0006823743,0.0001091726],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"qualitative","study_design_scores_codex":[0.0009940087,0.0002115572,0.09087586,0.00005768084,0.001063873,0.00005727264,0.7491104,0.001406459,0.01289839,0.000887457,0.0001424727,0.1422945],"study_design_scores_gemma":[0.001630131,0.00005152816,0.03773587,0.0004829759,0.001201429,0.000008006327,0.88789,0.06076718,0.004606321,0.005235161,0.0000271747,0.0003641736],"study_design_candidate":"qualitative","study_design_consensus":"qualitative","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9623452,0.003607712,0.0324846,0.0004789315,0.0001099963,0.000178867,0.0002380169,0.00005542424,0.0005012396],"genre_scores_gemma":[0.9949092,0.00004073475,0.003035218,0.0002263354,0.0001246176,0.000006277685,0.000730454,0.000009269813,0.0009179416],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1419304,"threshold_uncertainty_score":0.998786,"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."}}