© 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. 25-TRANSCODER-HP
    7. 108835
    Prev
    Next
    INDEX
    Explanations

    as

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    igrams
    -0.30
    syntax
    -0.28
    gings
    -0.28
    mtx
    -0.27
    abra
    -0.26
     mê
    -0.26
    gı
    -0.25
     Glock
    -0.25
    EDA
    -0.25
    gies
    -0.25
    POSITIVE LOGITS
    åį®
    0.29
    ,
    0.28
    åį¤
    0.27
     because
    0.26
    åľī
    0.26
    ãĥŀãĥ«
    0.26
     based
    0.25
     dependent
    0.25
    ?
    0.24
    usan
    0.24
    Activations Density 0.915%

    No Known Activations