Unverified50% confidenceFactExact time
SemantEq's design separates the generation of equivalence proofs from their acceptance, so that a stronger prover can raise recall without changing what acceptance means.
1
Sources
50%
Confidence
Long-term
Relevance
10/26/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe hierarchical verifier expands the evaluation result into multiple sub-scores, making the reward signal denser and the source of errors easier to locate, which is more advantageous in reinforcement learning optimization.69% similarUnverifiedSemantEq judges equivalence by checking proof terms rather than tactic scripts, to avoid bias introduced by differences in tactic style.69% similarUnverifiedSemantEq-Eval is a model-free search method that achieves 100% precision and 65.6% recall (based on human-adjudicated labels) on the ProofNetVerif test set.66% similarUnverifiedPerforming lossy but controllable compilation at the evidence level can actually improve the accuracy of the final answer.66% similarUnverifiedIn speculative decoding, the block length k should match the effective branching factor; too large a k amplifies the cost of mismatch.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1098300API
curl https://kongchang.com/api/v1/knowledge/claims/1098300MCP
get_claim(id=1098300)