© 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. 141904
    Prev
    Next
    INDEX
    Explanations

    conditions

    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
    ews
    -0.33
    ä¼ļå¼ķèµ·
    -0.33
    ä¼ļéĢłæĪIJ
    -0.33
    ä¼ļç»Ļ
    -0.31
    ä¼ļ导èĩ´
    -0.30
    ä¼ļ使
    -0.29
    åĵ¼
    -0.28
    åīĶ
    -0.27
    agrams
    -0.27
    ä¼ļ让
    -0.25
    POSITIVE LOGITS
    éŀį
    0.29
    ervation
    0.26
    )...
    0.25
    oir
    0.25
    uta
    0.25
    太æ¹ĸ
    0.23
    rei
    0.23
    dap
    0.23
    rates
    0.23
    juana
    0.23
    Activations Density 9.256%

    No Known Activations