Issues

• CharXiv reasoning selects instructions from question source instead of answer type (766/1000 validation prompts)

  EvolvingLMMs-Lab/LongVT#27 · Filter by author surgan12 opened on Oct 15, 2026

• Verify evals on Papers with Code

  EvolvingLMMs-Lab/LongVT#26 · Filter by author NielsRogge opened on Aug 16, 2026

• 无法复现 LongVT-7B-SFT 在 VideoMME 上的结果

  good first issuequestion
  Bug

  EvolvingLMMs-Lab/LongVT#25 · Filter by author Physicsmile opened on Jun 22, 2026
  2

  

• 请问在RL阶段，训练时模型每次的rollout都是实际调用了工具并且把结果返回给下一轮的推理吗？还是只模拟tool response，没有实际获取工具调用的实际结果呢？

  documentationgood first issue
  Feature

  EvolvingLMMs-Lab/LongVT#16 · Filter by author alanshaoTT opened on Feb 8, 2026
  6

  

• Question about metrics when evaluating LVBench

  good first issuehelp wantedquestion
  Task

  EvolvingLMMs-Lab/LongVT#11 · Filter by author Yusux opened on Dec 29, 2025
  8