© 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. 30-TRANSCODER-HP
    7. 95732
    Prev
    Next
    INDEX
    Explanations

    damage or harm

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    滥çĶ¨
    -0.28
     Exist
    -0.27
    ains
    -0.25
    åħįéĻ¤
    -0.25
     Exists
    -0.25
    SOR
    -0.25
    使çĶ¨æĿĥ
    -0.25
    ä¸į对
    -0.24
     SRC
    -0.24
    SRC
    -0.24
    POSITIVE LOGITS
     caused
    0.36
    éĥ½å·²ç»ı
    0.28
    éĥ½è¢«
    0.27
    éĢłæĪIJ
    0.27
    éĢłæĪIJçļĦ
    0.26
    桧
    0.26
    已被
    0.25
     already
    0.25
     Ca
    0.25
    éĽ£
    0.25
    Activations Density 0.002%

    No Known Activations