Unverified50% confidenceFactExact time
The blogger clarified that the claim that Astra improved the math ability of Qwen 3.5 0.8B by editing 0.016% of the weights was actually about locating and fixing a loop defect that prevented it from outputting a final answer, rather than injecting mathematical knowledge; the same operation had no effect on Gemma 3 270M, which has no loop defect.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/6/2026
First Seen
Valid until: 2/4/2027
Sources
LLMs Can Directly Read and Write Model Weights: A New Paradigm for Weight Space Programming
youtubeNeuralNine11/6/2026
Related Entities
Related Claims
UnverifiedIn experiment 8, Astra fixed the "looping" problem of the Qwen 3.5B (800 million parameter) model by editing its weights, after which the Math 500 score rose from slightly below 8% to slightly above 48% and GSM8K rose from 13.8% to 67.1%.78% similarUnverifiedBy editing an extremely small portion of the weights, Astra raised QUEN 3.5 0.8b's solve rate on math problems from roughly 1 correct out of every 13 problems to close to half.75% similarUnverifiedIn the experiment with the smallest Qwen 3.5 version (about 800 million weights), Astra modified 64 columns of numbers in each of layers 15 and 19 (about 130,000 weights), raising the standard math test score from under 8% to 48% and elementary word problems from 14% to 67%.72% similarUnverifiedAs a control, for the larger Gemma 3 model, which rarely loops, Astra could not improve its math score.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1143758API
curl https://kongchang.com/api/v1/knowledge/claims/1143758MCP
get_claim(id=1143758)