© 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. 18823
    Prev
    Next
    INDEX
    Explanations

    arguments/suggestions

    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
    èħ¥
    -0.28
    æŀī
    -0.25
     carc
    -0.25
    åŀ¢
    -0.25
    lit
    -0.24
     dynam
    -0.24
    æŶéļĶ
    -0.24
    idis
    -0.24
    失踪
    -0.24
    lined
    -0.23
    POSITIVE LOGITS
    airo
    0.31
    移
    0.30
    åIJijåīį
    0.30
     evolve
    0.29
    ç»ı
    0.29
    erset
    0.28
     move
    0.28
     shift
    0.27
    æݨ
    0.27
     moving
    0.27
    Activations Density 0.717%

    No Known Activations