© 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. 18963
    Prev
    Next
    INDEX
    Explanations

    other

    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
     but
    -0.63
    but
    -0.59
    ä½Ĩ
    -0.58
    ä½Ĩåľ¨
    -0.53
    ä½Ĩè¿Ļ
    -0.52
    ä½Ĩæĺ¯
    -0.49
    ä½ĨæĪij们
    -0.49
    _but
    -0.49
    But
    -0.48
    ä½ĨçĶ±äºİ
    -0.48
    POSITIVE LOGITS
    èĢĮæĺ¯
    0.90
     mÃł
    0.75
    èĢĮä¸Ķ
    0.63
    èĢĮ
    0.63
     utan
    0.59
     sino
    0.56
     instead
    0.56
     nor
    0.53
     بÙĦ
    0.50
     rather
    0.47
    Activations Density 1.399%

    No Known Activations