© 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. 20-TRANSCODER-HP
    7. 63623
    Prev
    Next
    INDEX
    Explanations

    zip code

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Wild
    -0.29
     Zucker
    -0.28
     Wild
    -0.26
    ç²¾éĢī
    -0.26
    ä¼ĺåĮĸ
    -0.26
     skipping
    -0.26
    aptor
    -0.26
    åħ¶ä»ĸåĽ½å®¶
    -0.25
     Occ
    -0.25
     optimization
    -0.25
    POSITIVE LOGITS
    /INFO
    0.29
    dress
    0.28
    udad
    0.26
    boro
    0.26
    ',['../
    0.25
    @app
    0.25
     hud
    0.24
    è·Ĺ
    0.24
    åľ°
    0.24
    trie
    0.24
    Activations Density 2.143%

    No Known Activations