© 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. 108394
    Prev
    Next
    INDEX
    Explanations

    Animals

    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
    elial
    -0.26
    å¯ħ
    -0.25
    ä½İ
    -0.25
     crackers
    -0.24
    ness
    -0.24
    (ele
    -0.24
    æıIJçĤ¼
    -0.24
    utive
    -0.24
    ARDS
    -0.23
    èĥ¶åĽĬ
    -0.23
    POSITIVE LOGITS
    èı±
    0.30
    ENTE
    0.28
     burst
    0.27
     spor
    0.25
    ä¹Ĺ
    0.25
    anz
    0.25
    åĬŀ好
    0.24
    çĥ¦
    0.24
    ENN
    0.24
     attachment
    0.24
    Activations Density 8.263%

    No Known Activations