LLMVis / tests /test_model_patterns.py

Commit History

  1. Bug fix: type issue with models leading to gibberish outputs

    5528a77

  1. feat(output): Add token scrubber with per-position top-5 next-token probabilities⁠

    cdpearlman +1: Cursor
    d60cfe2
  2. Added full sequence to analysis, still need to do output scrubber

    3f991b4

  1. Fix multi-layer ablation to ablate all heads simultaneously

    ae08976

  1. Add comprehensive test suite for utility modules (73 tests)

    e13b1ba