Unverified50% confidenceBenchmarkExact time
On the TriviaQA dataset, the learned answer selector improved average coverage from 20.2% to 30.9%.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOn TriviaQA, the error rate of answered questions (under the exact match scoring criterion) decreased from 23.0% to 14.5%.67% similarUnverifiedOn the news classification task, few-shot prompting achieved an average improvement of +24 percentage points; on the legal text task the improvement was only +3.4 percentage points, with two models showing performance degradation.67% similarUnverifiedOfficial data shows that after decision model training, a model's decision accuracy can increase from 30% to 80%.65% similarUnverifiedFiorillo v0.5 outputs probabilities for each candidate answer, addressing a three-way classification question on the Evidence Inference 2.0 benchmark: whether an intervention significantly increases, significantly decreases, or does not significantly change a given outcome measure relative to a control group.64% similarUnverified在歧义点的四类行为中，'先搜再问'成功率最高，平均通过率达93.4%，远高于'直接猜'的56.5%，'搜多次再猜'为51.5%64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1002058API
curl https://kongchang.com/api/v1/knowledge/claims/1002058MCP
get_claim(id=1002058)