Unverified50% confidenceBenchmarkExact time
SemantEq-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.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/26/2026
First Seen
Valid until: 1/24/2027
Sources
Related Entities
Related Claims
UnverifiedIFEval基准测试专门评估模型遵循可验证指令的能力，测试显示即使最强闭源模型在严格约束条件下的遵循率也难以达到90%以上67% similarUnverifiedSemantEq's design separates the generation of equivalence proofs from their acceptance, so that a stronger prover can raise recall without changing what acceptance means.66% similarUnverifiedAn independent reproduction test ran the full benchmark using the project's own executor, and the results matched BENCHMARK.md byte for byte, confirming that the 98% figure itself is not fabricated.66% similarUnverifiedEngy's verification is sampling-based and probabilistic, not a complete proof of every forward pass, and the verification only proves which model was run without providing privacy guarantees.65% similarUnverifiedSemantEq judges equivalence by checking proof terms rather than tactic scripts, to avoid bias introduced by differences in tactic style.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1098294API
curl https://kongchang.com/api/v1/knowledge/claims/1098294MCP
get_claim(id=1098294)