© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 31-TRANSCODER-HP
    7. 19650
    Prev
    Next
    INDEX
    Explanations

    the

    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
     freely
    -0.30
    è·¡
    -0.27
    utely
    -0.25
     bars
    -0.25
    åįĩéĻį
    -0.25
     more
    -0.25
    åŃĹæł·
    -0.25
    ç»ĦåĽ¾
    -0.25
    Smarty
    -0.24
    åħ·ä½ĵæĥħåĨµ
    -0.24
    POSITIVE LOGITS
    following
    0.35
     following
    0.32
    door
    0.31
    以ä¸ĭ
    0.30
     door
    0.29
    _follow
    0.29
    _that
    0.28
    以ä¸ĭãģ®
    0.28
     Ñģлед
    0.28
     unin
    0.28
    Activations Density 0.373%

    No Known Activations