Unverified50% confidenceBenchmarkExact time
According to Table 3 of the paper, Gemma-4-12B-it achieves a single-step accuracy of 74.4% and an end-to-end task success rate of only 5.0%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/5/2026
First Seen
Valid until: 2/3/2027
Sources
SLATE Benchmark Exposes AI Agent Weakness: 74% Single-Step Accuracy but Only 5% Task Success Rate
redditr/LangChain11/5/2026
Related Entities
Related Claims
UnverifiedThe simple 3-rule baseline used for comparison achieved an accuracy of only 64.4%, with Jev leading by approximately 8 percentage points.73% similarUnverifiedBullet 在 SWE-bench Verified 基准测试中取得 95.8% 的成绩，位列前三，平均每个任务耗时 119 秒72% similarUnverifiedNVFP4量化方案相比BF16基线实现了1.5倍的推理加速，同时在top-1准确率上保留了92%到97%的水平71% similarUnverifiedGazeEval achieves 71.3% accuracy on the prediction task.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1138672API
curl https://kongchang.com/api/v1/knowledge/claims/1138672MCP
get_claim(id=1138672)