Unverified50% confidenceBenchmarkExact time
Without any training, Astra modified only 0.016% of the weights of a small model, raising its score on the Math 500 benchmark from 7.9% to 48.1%.
1
Sources
50%
Confidence
Long-term
Relevance
11/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn a math repair experiment, a small open-source model scored only 7.9% on the Math 500 test; Astra raised the score to 48.1% by locating and subtracting a direction within two layers (about 130,000 weights, an edit ratio of 0.016%), and its instruction-following ability also improved roughly twofold.80% similarUnverifiedIn the experiment, Astra modified the weights of the small Qwen model, raising the MATH500 score from 7.9% to 48.1%, while the share of answers hitting the trial-run limit fell from 93% to 33%.78% similarUnverifiedAt the 1B scale, AsymLoop's average zero-shot accuracy on seven common-sense reasoning benchmarks improves from 49.58% to 50.85% (compared with Full Loop).74% similarUnverifiedWhen sampling eight candidates without constraints from the model, 10.2% of inputs cannot produce any valid output under limited training data, and this proportion drops to 0.4% when using the full training data71% similarVerifiedWhen fine-tuning with LoRA, the actual number of parameters that need to be trained is only 0.1%-1% of the original.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1135380API
curl https://kongchang.com/api/v1/knowledge/claims/1135380MCP
get_claim(id=1135380)