Unverified50% confidenceFactExact time
GFlowNets sample proportionally rather than maximizing reward, so even when the out-of-sample accuracy of the reward proxy is low, the model maintains broad exploration and is robust to an imperfect reward proxy.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWhen the success rate is extremely low, almost all sample rewards are identical, resulting in extremely high policy gradient variance and very low signal-to-noise ratio, preventing the model from obtaining effective update directions.74% similarUnverifiedBlind testing by domain experts found that CollabSciReM is significantly superior to general reward variants on novelty and reports fewer hallucinated responses.70% similarUnverifiedWhen the within-group reward distribution is severely skewed, the mean is pulled off by extreme values, causing distorted advantage estimates; in such cases, the median or a trimmed mean may be more robust.67% similarUnverifiedIn the synthetic experiment, "always output RISK_LEVEL: 1" achieves an ordinary accuracy as high as 0.922, but an average reward of −0.000, showing that class-balanced rewards can prevent laziness.65% similarUnverifiedConsistency of output quality is often difficult to reflect in benchmark scores; a model with a high average score but occasional severe errors carries higher risk than a stable but slightly weaker model in certain business scenarios.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1113771API
curl https://kongchang.com/api/v1/knowledge/claims/1113771MCP
get_claim(id=1113771)