© 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. 87687
    Prev
    Next
    INDEX
    Explanations

    Mixed text and code

    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
    è·Ŀ离
    -0.28
    æľ¬çĿĢ
    -0.28
     distance
    -0.28
    è¾¹éĻħ
    -0.26
    åįĬå¤ľ
    -0.25
    ctal
    -0.24
    loff
    -0.24
    é£ŁåłĤ
    -0.24
    rade
    -0.24
    åħ¬äº¤
    -0.23
    POSITIVE LOGITS
    /msg
    0.29
    åĴĮå¹³
    0.28
     sweetheart
    0.25
    示æĦı
    0.24
    è§ģæķĪ
    0.24
    ç»ĵæĻ¶
    0.24
    鸣
    0.24
    çĻ½çĻ½
    0.24
    Intern
    0.24
    æĽ°
    0.23
    Activations Density 0.002%

    No Known Activations