© 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. 35-TRANSCODER-HP
    7. 40092
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    Not in Any Lists

    No Comments

    Negative Logits
    çŃĶ
    -0.29
    igkeit
    -0.28
    swer
    -0.27
    -Mail
    -0.27
    uld
    -0.27
    éĿ¢
    -0.26
    =color
    -0.25
    /vue
    -0.25
    yme
    -0.25
    uter
    -0.25
    POSITIVE LOGITS
     populated
    0.26
    iki
    0.26
    ä¸ļåĬ¡
    0.25
    missions
    0.25
    (boost
    0.24
    å«ĸ
    0.23
    åĮºåĪĨ
    0.23
    -death
    0.23
    _normalized
    0.23
    MLElement
    0.23
    Activations Density 1.710%

    No Known Activations

    This feature has no known activations.