Unverified50% confidenceFactExact time
Standard GRPO samples multiple trajectories from the task's initial state, whereas ReTerm-GRPO changes the sampling starting point to states in which the agent itself has already failed.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe core innovation of GRPO is the complete removal of the value network; it samples multiple responses to the same question to form a group, and uses the mean return within the group as the baseline.67% similarUnverifiedWhen Grep and Glob were unable to start a search due to exhausted system resources, they previously incorrectly reported no matches; they now return a clear error message.66% similarUnverifiedgrep can only return matching lines of text and cannot determine which results are truly valuable for the current task.65% similarUnverifiedThe key innovation of GRPO compared to the traditional PPO algorithm is that it replaces the independent value network by sampling multiple outputs for the same problem and computing relative scores within the group, thereby reducing memory usage.65% similarUnverifiedGRPO generates a group of candidate outputs for the same input and updates the model using intra-group relative reward signals rather than absolute reward signals, reducing reliance on an independent value network (critic).62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1023940API
curl https://kongchang.com/api/v1/knowledge/claims/1023940MCP
get_claim(id=1023940)