Unverified50% confidenceFactExact time
The Token Entanglement Hypothesis holds that the root cause of the unconscious number phenomenon lies in tokens of different semantic categories in the model's output layer (Unembedding Matrix) being close to each other in high-dimensional vector space, causing activation to leak into adjacent categories.
1
Sources
50%
Confidence
Long-term
Relevance
9/18/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedFlattening a matrix and feeding it to an MLP results in a non-fixed input dimension due to variable row counts, and breaks permutation invariance.67% similarUnverifiedContext window inflation caused by intermediate tokens or conflicts with compression mechanisms are an entirely different class of problems from the semantic validity of reasoning tokens.66% similarUnverified仅无约束地最大化SWD会导致神经网络扭曲数据结构得到退化解，因此需引入几何约束在分离性与保真度之间权衡64% similarUnverifiedToken compression involves trade-offs: some seemingly redundant contextual information may be critical for model reasoning, and aggressive compression may affect output quality in scenarios such as code generation and complex reasoning.64% similarUnverified模型规模越大，幻觉的表述往往越流畅、越具迷惑性，识别难度反而上升63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/931310API
curl https://kongchang.com/api/v1/knowledge/claims/931310MCP
get_claim(id=931310)