Unverified50% confidenceFactExact time
Reward routing refers to assigning training data to different processing paths based on sample scores under the reward model, to prevent low-quality samples from contaminating training.
1
Sources
50%
Confidence
Long-term
Relevance
11/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedExamining the model's action sequences during training one by one can determine whether high scores result from genuine capability improvement or from exploiting loopholes in the reward mechanism.79% similarUnverifiedSatisDive adopts a training-free strategy that takes effect at inference time, introducing batch-relative reward truncation to distinguish between low-reward and high-reward candidates.77% similarVerifiedIn reinforcement learning training, the model's goal is to maximize the score given by the reward function.76% similarUnverifiedThe advantage of guidance during sampling is its low computational overhead, the flexibility to switch rewards, and that it does not destroy the natural sequence distribution learned in pretraining; the disadvantage is that the guidance strength is difficult to control finely and tends to globally distort the generation results.75% similarUnverifiedIn such tasks, reinforcement learning usually means having the model generate multiple candidate answers to the same question, then reinforcing high-scoring outputs and suppressing low-scoring ones based on a reward signal; crude reward design can induce score-gaming behavior.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1174207API
curl https://kongchang.com/api/v1/knowledge/claims/1174207MCP
get_claim(id=1174207)