© 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. 18048
    Prev
    Next
    INDEX
    Explanations

    Survey statistics

    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
    _require
    -0.28
    éħ¥
    -0.25
     imprint
    -0.25
    ør
    -0.25
    ower
    -0.24
     birds
    -0.24
    prene
    -0.24
    langs
    -0.24
    ext
    -0.23
     onSubmit
    -0.23
    POSITIVE LOGITS
     seen
    0.32
     opinion
    0.31
    è°ĥæŁ¥
    0.31
    æĶ¯æĮģ
    0.31
     dis
    0.31
     survey
    0.30
     strongly
    0.30
    survey
    0.30
    çľĭåĪ°
    0.30
    seen
    0.30
    Activations Density 0.177%

    No Known Activations