© 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. 30-TRANSCODER-HP
    7. 95499
    Prev
    Next
    INDEX
    Explanations

    Scraped text snippets

    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
    Not in Any Lists

    No Comments

    Negative Logits
    stances
    -0.28
    ãĤĬ
    -0.27
    ames
    -0.27
    çŀ½
    -0.26
    ç³Ĭæ¶Ĥ
    -0.25
    大éĻ¸
    -0.25
    两岸
    -0.25
    omorphic
    -0.24
    .prop
    -0.24
    ettle
    -0.24
    POSITIVE LOGITS
    é¢Ŀ
    0.28
    ç¬Ķ
    0.27
    éĽĨ群
    0.26
    æ´»åĬ¨
    0.26
     caused
    0.25
    åŁºåľ°
    0.25
    scope
    0.24
    è°ĥèĬĤ
    0.24
    io
    0.23
     writing
    0.23
    Activations Density 0.884%

    No Known Activations