© 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. 32-TRANSCODER-HP
    7. 43254
    Prev
    Next
    INDEX
    Explanations

    code/research

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ç²ī
    -0.29
    æ¶ĵ
    -0.28
    å¾®
    -0.27
    pow
    -0.27
    æĢ¡
    -0.26
     Sie
    -0.26
    Sie
    -0.26
    è´¢
    -0.26
    ä¸ĭéĿ¢æĺ¯
    -0.25
    å¸Į
    -0.24
    POSITIVE LOGITS
     research
    0.36
    Research
    0.34
     Research
    0.32
    research
    0.30
    osome
    0.27
     landscape
    0.26
     bases
    0.26
     researching
    0.25
    ammer
    0.25
     preservation
    0.25
    Activations Density 0.001%

    No Known Activations