© 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. 35-TRANSCODER-HP
    7. 38723
    Prev
    Next
    INDEX
    Explanations

    build

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    URA
    -0.30
    urat
    -0.29
    apter
    -0.28
    atile
    -0.27
    BX
    -0.26
    aira
    -0.26
    anova
    -0.25
    å»ĸ
    -0.25
    ales
    -0.25
    avit
    -0.25
    POSITIVE LOGITS
    oder
    0.35
    odor
    0.35
    od
    0.29
    odies
    0.29
    odate
    0.28
    oday
    0.28
    odo
    0.27
    ä»İæѤ
    0.27
    odu
    0.26
    åįģ大
    0.26
    Activations Density 0.904%

    No Known Activations