Unverified50% confidenceBenchmarkExact time
On held-out Qwen3-32B dialogue segments, with the threshold set to flag 5% of benign calls, the linear probe caught 82% of leaking calls.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedQwen3 27B在OpenAI HumanEval 164道Python题上取得85%通过率和99%应答率73% similarUnverifiedThe random-position null baseline for Qwen3-4B was only 11.3%, and the upper bound for directly patching lower-level prompt facts was 83%.73% similarUnverifiedUsing Qwen3-14B to generate risk exam papers yielded only 5/25 passages with clear signals, and only after switching to stricter prompts and Qwen3-32B were 21 passages obtained.72% similarUnverifiedWhen run on Qwen3.6-35B-A3B with a 50% expert pruning ratio, SlimWise increases decoding throughput by up to 1.81x with minimal accuracy loss.72% similarUnverifiedIn agentgateway's testing, qwen3.8 (27.3B, Q4_K_M quantization, 262K context) had a valid answer rate of 90.0% (54/60), an accuracy rate among answered questions of 98.1% (53/54), and an overall accuracy rate of 88.3% (53/60).72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1115383API
curl https://kongchang.com/api/v1/knowledge/claims/1115383MCP
get_claim(id=1115383)