© 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. 33-TRANSCODER-HP
    7. 54222
    Prev
    Next
    INDEX
    Explanations

    musician

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    bon
    -0.32
    ol
    -0.27
    沸
    -0.27
    verts
    -0.26
    bra
    -0.26
    .Unsupported
    -0.26
    aint
    -0.26
    ÑĢÑĥг
    -0.25
    bol
    -0.25
     bra
    -0.25
    POSITIVE LOGITS
     ká»·
    0.27
    åµĮ
    0.26
    ç§ĥ
    0.26
    æīĵæ³ķ
    0.25
     SME
    0.25
    moth
    0.24
    imen
    0.23
    inue
    0.23
    å´ģ
    0.23
    /gin
    0.23
    Activations Density 0.009%

    No Known Activations