© 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. 22-TRANSCODER-HP
    7. 112094
    Prev
    Next
    INDEX
    Explanations

    Mathematical expressions and numbers

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    )['
    -0.31
     llev
    -0.29
     assignable
    -0.28
    )["
    -0.28
    åı¯è§Ĥ
    -0.27
    ,['
    -0.27
    leston
    -0.26
    ijd
    -0.26
    raisal
    -0.25
    etsk
    -0.25
    POSITIVE LOGITS
     nh
    0.26
    perimental
    0.24
    åıijå±ķè¶ĭåĬ¿
    0.24
    èĬ
    0.24
    cade
    0.23
    è³ĵ
    0.23
    对åħ¶çľŁå®ŀ
    0.23
    åŃĹ第
    0.23
    åĴ¯
    0.23
     bulls
    0.23
    Activations Density 0.008%

    No Known Activations