© 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. 30-TRANSCODER-HP
    7. 96940
    Prev
    Next
    INDEX
    Explanations

    special

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    oom
    -0.32
    วà¸ĩ
    -0.29
    åĭī
    -0.26
    è¿IJåĬ¨
    -0.26
    bol
    -0.25
    岩
    -0.25
    å¨ĺ
    -0.25
    pen
    -0.25
    esper
    -0.24
    ursed
    -0.24
    POSITIVE LOGITS
    注æĺİåĩºå¤Ħ
    0.28
    aptured
    0.28
     ÄijÃŃch
    0.27
    ampo
    0.26
    (::
    0.26
    [at
    0.25
    ifacts
    0.25
    æĶ¶èĹı
    0.25
     Provid
    0.25
     Holding
    0.25
    Activations Density 0.001%

    No Known Activations