Unverified50% confidenceSolutionExact time
Hard negative samples (audio similar to the target keyword but that should not trigger, such as homophones and similar phrases) can force the model to focus on truly discriminative features, thereby reducing false activations in real environments, and are more effective than simply increasing the amount of positive samples.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedStreaming models have more failure modes than batch models and often hallucinate additional speakers out of thin air, which is a direct consequence of insufficient context.78% similarUnverifiedHigh bias means the model is too simple and leads to underfitting, while high variance means the model has learned noise and leads to overfitting, and the two tend to trade off against each other.77% similarVerifiedNoise information in the context degrades model reasoning accuracy in a nonlinear manner77% similarUnverifiedRude and hostile prompts tend to degrade the quality of model output, and the model may give shorter, uncooperative, or otherwise lower-quality responses.77% similarUnverifiedThe study proposes using a causal counterfactual criterion to suppress false starts: if muting the user input leaves the model's output distribution nearly unchanged, the event is identified as a false start and suppressed.77% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1010406API
curl https://kongchang.com/api/v1/knowledge/claims/1010406MCP
get_claim(id=1010406)