Unverified60% confidenceSolutionExact time
COCONUT (Chain of Continuous Thought) propagates the reasoning chain directly in continuous latent space, feeding the hidden state produced by the previous reasoning step directly as input to the next, bypassing the encode-decode loop from hidden state to token and back to hidden state.
2
Sources
60%
Confidence
Long-term
Relevance
9/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedHidden Chains of Thought do not require explicitly displaying each step of the thinking process in the output, allowing the model to perform reasoning iterations within its internal state space.77% similarUnverifiedThe Coconut (Chain of Continuous Thought) research proposed by the MIT team allows models to perform multi-step reasoning in a continuous latent space without discretizing intermediate states into text.75% similarUnverifiedThe Preserved Thinking mechanism avoids KV Cache invalidation and reasoning state loss caused by tool calls by fixing previous reasoning chains in a structured form within the context prefix region.70% similarUnverifiedLatent space reasoning allows the model to complete thinking in a continuous vector space composed of hidden states, only decoding the result into text at the final stage.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/852734API
curl https://kongchang.com/api/v1/knowledge/claims/852734MCP
get_claim(id=852734)