Unverified50% confidenceBenchmarkExact time
The Mindscend team's 26.2M-parameter GECToR-style edit annotator v1.5 achieves 0.799 on the F0.5 metric, with a 2.4% false-edit rate and a 0.0% protected-span loss rate.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/12/2026
First Seen
Valid until: 1/10/2027
Sources
Related Entities
Related Claims
UnverifiedWith two to three attempts, GLM-5.3 Flash scored 82%, which is within the margin of error of GPT-5.6 Luna's 81%.67% similarUnverifiedThe same parameter name reasoning_effort behaves differently across models from different vendors: in gpt-5.4-mini the median output tokens rose from 18 at none to 254 at high (×14.1); gpt-oss-120b rose from refusal to 1,340 (×4.6); claude-haiku-5.5 rose from 150 to 298 (×2.0); gemini-3.7-flash rose from 453 to 1,013 (×2.2).67% similarUnverifiedGAD-RL achieved an Overall total score of 91.18 on OmniDocBench v1.666% similarUnverifiedThe C runtime's perplexity is 26.8407, and the PyTorch reference implementation's is 26.8188, a difference of 0.082%, meeting the threshold of <=1%.66% similarUnverifiedThe parsing defect combined with one incorrect answer key caused gpt-oss's score across five core tasks to shift from 0.62 to 1.00, a shift as large as 0.38.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1009449API
curl https://kongchang.com/api/v1/knowledge/claims/1009449MCP
get_claim(id=1009449)