© 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. 19614
    Prev
    Next
    INDEX
    Explanations

    gibberish

    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
    亲
    -0.29
     wake
    -0.26
    ç¡®
    -0.25
     TASK
    -0.25
    ç«ĭåĪ»
    -0.25
     wast
    -0.25
    å¿ħå°Ĩ
    -0.24
     suck
    -0.24
    ä¸ĭåİ»
    -0.24
    认
    -0.24
    POSITIVE LOGITS
    uter
    0.25
    åĩºåħ·
    0.25
    è¯ķ管婴åĦ¿
    0.24
     terror
    0.24
    辦çIJĨ
    0.24
    åĪ¶è£ģ
    0.23
    {name
    0.23
    oidal
    0.23
     {}:
    0.23
    <_
    0.23
    Activations Density 4.046%

    No Known Activations