© 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. 31-TRANSCODER-HP
    7. 19031
    Prev
    Next
    INDEX
    Explanations

    Mary

    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
    æ±°
    -0.31
    çĥĤ
    -0.29
    hits
    -0.28
    ipop
    -0.27
     DEAL
    -0.27
    istol
    -0.27
    awl
    -0.26
    achat
    -0.26
     Hits
    -0.26
    ibal
    -0.26
    POSITIVE LOGITS
     direct
    0.34
    Direct
    0.32
    ux
    0.31
     complement
    0.30
    zer
    0.29
    æĿ¾
    0.29
     indirect
    0.29
    éĹ´æİ¥
    0.29
    åĪĨ
    0.28
    æĿ¾å¼Ľ
    0.28
    Activations Density 0.016%

    No Known Activations