Search results

Data Quality Concerns: Conflicting System Prompts and Poor Context Formatting

Status: Open.

#15 In Tencent-Hunyuan/CL-bench;

· wintertee opened on Mar 6, 2026

GPT-5.2 scores seem off

Status: Open.

#13 In Tencent-Hunyuan/CL-bench;

· xeophon opened on Mar 4, 2026

How to Reproduce Error Type Distribution in Table 3?

Status: Open.

#12 In Tencent-Hunyuan/CL-bench;

· Flipped-ciks opened on Feb 27, 2026

CL-Bench Usage

Status: Open.

#8 In Tencent-Hunyuan/CL-bench;

· LLMforScience opened on Feb 9, 2026

Request: Broader model coverage (Mistral, xAI/Grok, Cerebras, Groq-hosted)

Status: Open.

#7 In Tencent-Hunyuan/CL-bench;

· clemlesne opened on Feb 7, 2026

Improve messages field in CL-bench by adding a delimiter to separate context and task.

Status: Open.

#3 In Tencent-Hunyuan/CL-bench;

· 9sxx opened on Feb 4, 2026