© 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. 18283
    Prev
    Next
    INDEX
    Explanations

    News, arguments, imperatives

    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
    oref
    -0.28
     Need
    -0.26
    -floating
    -0.26
    theless
    -0.25
    纪å½ķ
    -0.25
     vests
    -0.24
     testimon
    -0.24
     needing
    -0.24
    å«Įçĸij
    -0.24
     pits
    -0.24
    POSITIVE LOGITS
     if
    0.43
     not
    0.41
     right
    0.35
     across
    0.34
    å¦ĤæŀľæĪij们
    0.33
    if
    0.33
     fast
    0.33
     forward
    0.32
    ours
    0.31
    å¦Ĥæŀľ
    0.31
    Activations Density 2.308%

    No Known Activations