Unverified50% confidenceBenchmarkExact time
In a benchmark on 50 synthetic chore scenarios, the fine-tuned SlothDistill, compared with the base Llama-3-8B, reduced p95 decision latency from 1,420 ms to 185 ms (a 7.6x improvement), cut average output tokens from 412 to 48 (an 88.3% saving), raised direct-action accuracy from 62.4% to 97.8%, and lowered API cost per thousand tasks from $1.85 to $0.22 (88.1% cheaper).
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/20/2026
First Seen
Valid until: 1/18/2027
Sources
Related Entities
Related Claims
UnverifiedIn the dual-bug Python function repair benchmark (215 hidden tests), omitting effort (max) consumes an average of 6,364 reasoning tokens, takes 56 seconds, and costs $0.0045 per run; effort=low consumes an average of 591 tokens, takes 9.5 seconds, and costs $0.0006 per run.71% similarUnverifiedDoubling the thinking token budget from 8k to 16k brought only a 0.2% accuracy improvement, while changing the distribution of failed questions.70% similarUnverifiedgpt-oss-120b spent about 25 times the cost per correct answer on the distractor counting task, while its accuracy was 7.5 percentage points lower (not significant).70% similarUnverifiedIn the 5-agent conversational cluster architecture (unstructured Markdown handoffs), the spec drift failure rate is 48.4%, the silent tool error propagation rate is 79.0%, the premature termination rate is 34.2%, and the end-to-end success rate is 18.5%.70% similarUnverified研究数据显示，在复杂代码生成任务中，将thinking budget从1024 tokens提升到8192 tokens可以带来15-30%的准确率提升，但推理成本增加4-8倍70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1072942API
curl https://kongchang.com/api/v1/knowledge/claims/1072942MCP
get_claim(id=1072942)