© 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. 24-TRANSCODER-HP
    7. 150105
    Prev
    Next
    INDEX
    Explanations

    Comparisons and qualifications

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_24.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.24.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     favoured
    -0.30
    å·´åŁºæĸ¯åĿ¦
    -0.29
     criticised
    -0.27
    åij¨ä¸ī
    -0.26
    åıģ
    -0.25
    bery
    -0.25
    ä¸įåĨįæĺ¯
    -0.24
    éĥ½ä¸įæĺ¯
    -0.24
    Jam
    -0.24
    ç¨
    -0.24
    POSITIVE LOGITS
    educt
    0.28
    æŁ¥
    0.27
    缩
    0.24
     summon
    0.24
     append
    0.24
     something
    0.24
    ä¸ĭçļĦ
    0.24
    åĩºç§Ł
    0.23
     administering
    0.23
     making
    0.23
    Activations Density 0.927%

    No Known Activations