© 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. 23-TRANSCODER-HP
    7. 23871
    Prev
    Next
    INDEX
    Explanations

    Inspection and analysis reports

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_23.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.23.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Alley
    -0.28
     Ever
    -0.26
     Paran
    -0.26
    .easy
    -0.25
    è¯Ŀ说
    -0.24
    ono
    -0.24
    coe
    -0.23
    头çļ®
    -0.23
    èģĮ
    -0.23
    Ëĥ
    -0.23
    POSITIVE LOGITS
     compared
    0.35
     carefully
    0.33
    ä»Ķç»Ĩ
    0.31
    -ground
    0.29
    intern
    0.28
    详ç»Ĩ
    0.28
     objectively
    0.27
    详ç»ĨäºĨ解
    0.27
     detail
    0.27
     once
    0.26
    Activations Density 0.004%

    No Known Activations