© 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. 19397
    Prev
    Next
    INDEX
    Explanations

    work

    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
     working
    -1.36
    å·¥ä½ľ
    -1.31
     work
    -1.23
    working
    -1.19
     Working
    -1.16
    Working
    -1.16
     worked
    -1.13
    çļĦå·¥ä½ľ
    -1.07
    -working
    -1.05
    _working
    -1.03
    POSITIVE LOGITS
    æŃĨ
    0.27
    bigint
    0.25
    æĹ©ãģĦ
    0.25
    uzzi
    0.25
    cheduler
    0.24
     Grill
    0.23
     grill
    0.23
    itech
    0.23
     shield
    0.22
    管çIJĨåĪ¶åº¦
    0.22
    Activations Density 29.684%

    No Known Activations