Unverified50% confidenceBenchmarkExact time
On the Reasoning Gym benchmark, using the Gemma-3-4B model, CRoS achieves 46.8% accuracy on five unseen logic and game datasets, surpassing the strongest baseline (43.8%).
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/2/2026
First Seen
Valid until: 1/31/2027
Sources
Related Entities
Related Claims
UnverifiedThe fine-tuned Qwen3.5-4B model improved accuracy on BU Bench V1 from 15% to 53%.70% similarUnverifiedOn 1,000 validation cases, the strongest LLM baseline achieves 62.7% accuracy, while EMBER (73.6%) and ORBIT (70.9%) perform better.67% similarUnverifiedIn the LongMemEval benchmark, BM25 accuracy improved from 78.8% to 85.2% with 5 chunks.67% similarUnverifiedOn Llama-3.2-3B, fine-tuning a brand-new dense model with the 5% of data selected by CACTUS reached 49.44% accuracy on BIG-Bench Hard; selecting data directly with the dense model gave 49.54%; data selected by the ALPS method could only reach 47.22%.67% similarUnverifiedThe simple 3-rule baseline used for comparison achieved an accuracy of only 64.4%, with Jev leading by approximately 8 percentage points.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1125786API
curl https://kongchang.com/api/v1/knowledge/claims/1125786MCP
get_claim(id=1125786)