© 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. 95849
    Prev
    Next
    INDEX
    Explanations

    have has

    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
    ka
    -0.28
    stalk
    -0.25
    åĨ³è®®
    -0.25
    маÑĢ
    -0.25
    åķ¸
    -0.25
     composing
    -0.24
    æ¼Ĥ
    -0.24
    Inspectable
    -0.23
    lex
    -0.23
     caus
    -0.23
    POSITIVE LOGITS
     needs
    0.35
    needs
    0.31
     fears
    0.30
     faculties
    0.28
     secrets
    0.28
     wants
    0.28
     feelings
    0.28
     unique
    0.27
     preferences
    0.27
    preferences
    0.27
    Activations Density 0.106%

    No Known Activations