© 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. 95941
    Prev
    Next
    INDEX
    Explanations

    that, which

    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
     Performing
    -0.28
     worn
    -0.28
     performing
    -0.28
     lat
    -0.26
    æī¶è´«å·¥ä½ľ
    -0.26
     performed
    -0.26
    Perform
    -0.25
    perform
    -0.25
    own
    -0.25
     coming
    -0.25
    POSITIVE LOGITS
     uses
    0.36
    èģļçĦ¦
    0.35
     focuses
    0.34
     foc
    0.34
     examines
    0.34
     examine
    0.33
    focus
    0.33
     focus
    0.33
     use
    0.32
     analyze
    0.29
    Activations Density 0.911%

    No Known Activations