New paper argues for a quasi-symbolic view: neural nets implement symbols only approximately
neuranna · x · 2026-10-14
Tom McCoy and colleagues publish a paper in Current Opinion in Behavioral Sciences arguing, based on mechanistic interpretability evidence, for a quasi-symbolic view of cognition.
- Neural network vector representations often show emergent symbolic structure: they decompose into meaningful constituents (subject/verb/object) and align with symbolic dimensions in vector space.
- Yet these vectors are only quasi-symbolic: they show graded similarity ("movie" is closer to "novel" than "meal") and graded context effects ("good" affects different nouns differently).
- Conclusion: neural nets implement symbols, but only approximately.
More from Research
- RFD4-Proteina's 'underrated' docs detail partial diffusion for protein redesign — AllThingsApx · 2026-10-23
- MPI-Sec PhD Thesis on Vulnerability Detection Benchmarks Includes Pre-LLM Classic "Top Score on the Wrong Exam" — mboehme_ · 2026-10-23
- LiveNerf's 29-day data finds no nerfing of Claude Opus 5.5 — TheOnlyVibemaster · 2026-10-23
- New paper makes 9 falsifiable predictions toward a major AI-enabled terror attack by 2027-2028 — StephenLCasper · 2026-10-23
- HumanoidPose recovers humanoid robots' full-body joint states from RGB video alone — cholz · 2026-10-23
- Sophont's Holon-S1 predicts genes and biomarkers from routine pathology slides — iScienceLuvr · 2026-10-23