© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 31-TRANSCODER-HP
    7. 18374
    Prev
    Next
    INDEX
    Explanations

    a/at

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    conditionally
    -0.32
    lessly
    -0.29
     plainly
    -0.27
     distinctly
    -0.27
     respectfully
    -0.26
    thesize
    -0.26
     clearly
    -0.25
    æ´Ĺæīĭ
    -0.24
     itself
    -0.24
    urally
    -0.24
    POSITIVE LOGITS
     quick
    0.31
     quickest
    0.28
    æĸij
    0.27
     rapid
    0.27
     Annex
    0.26
    å¿«éĢŁ
    0.26
     Quick
    0.26
    æīĵæĸŃ
    0.24
    quick
    0.24
    å½ĵä¹ĭ
    0.24
    Activations Density 0.006%

    No Known Activations