© 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. 18372
    Prev
    Next
    INDEX
    Explanations

    ],

    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
    atus
    -0.30
     versa
    -0.29
     fines
    -0.27
    culos
    -0.27
    èIJ½
    -0.27
     roll
    -0.26
    ill
    -0.26
     theirs
    -0.25
     rollers
    -0.25
    lickr
    -0.25
    POSITIVE LOGITS
    åıĮçľ¼
    0.31
    èĢ³
    0.27
    mith
    0.27
    åħį
    0.27
    åIJ¬äºĨ
    0.26
    æģ©
    0.26
    æĮĩ导
    0.26
    mium
    0.26
    ç®±
    0.25
    oste
    0.25
    Activations Density 0.073%

    No Known Activations