© 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. 111627
    Prev
    Next
    INDEX
    Explanations

    Code and/or math

    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
    åħ¶æ¬¡
    -0.28
    åıªçŁ¥éģĵ
    -0.27
    人们
    -0.26
    edition
    -0.25
    ä¼łè¯´
    -0.25
    -bo
    -0.24
    çīĪ
    -0.24
    åĦĴ
    -0.24
    åħ¶æ¬¡æĺ¯
    -0.24
    Else
    -0.24
    POSITIVE LOGITS
    iciary
    0.28
    (cps
    0.27
    @↵↵
    0.27
    otec
    0.26
    ldata
    0.26
    Reached
    0.25
    ickness
    0.25
    undles
    0.25
    uese
    0.24
     ×Ķ×ĸ
    0.24
    Activations Density 0.008%

    No Known Activations