© 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. 98863
    Prev
    Next
    INDEX
    Explanations

    Code and typos

    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
    æ´»çĿĢ
    -0.30
     haul
    -0.27
    .jetbrains
    -0.26
     STAR
    -0.25
    äºĮèĢħ
    -0.25
    ï½ľ
    -0.25
    }|
    -0.24
    URA
    -0.24
    -------------↵
    -0.24
    ITIONS
    -0.24
    POSITIVE LOGITS
    èĽŃ
    0.26
    ually
    0.24
    rength
    0.24
    ãģ¾ãģĦ
    0.23
    æľĪåĪĿ
    0.23
    ¶Ī
    0.23
    æ¸ħé¦Ļ
    0.23
    æ¶Īè²»
    0.23
    æĦıæĢĿ
    0.23
    olest
    0.23
    Activations Density 0.038%

    No Known Activations