© 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. 19963
    Prev
    Next
    INDEX
    Explanations

    formal titles and expressions

    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
    Negative Logits
    åĤ¨èĥ½
    -0.27
    賦
    -0.26
    èµĭ
    -0.25
    èµĭäºĪ
    -0.25
     newRow
    -0.25
     stem
    -0.25
    Steel
    -0.25
    igor
    -0.24
    mue
    -0.24
    oup
    -0.24
    POSITIVE LOGITS
    lush
    0.29
    :normal
    0.28
    轸
    0.28
    èĹ©
    0.28
    ungi
    0.27
    åľ°æłĩ
    0.27
    è§ĤæľĽ
    0.25
    å°º
    0.25
    _land
    0.25
    åľŁåľ°
    0.24
    Activations Density 0.006%

    No Known Activations