© 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. 25-TRANSCODER-HP
    7. 107661
    Prev
    Next
    INDEX
    Explanations

    for

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.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¿ĻäºĽä¸ľè¥¿
    -0.29
    tlement
    -0.27
     Returned
    -0.26
    pairs
    -0.25
    iid
    -0.25
     understandably
    -0.25
    å§ļ
    -0.25
     Tabs
    -0.25
    å¼Ħå¾Ĺ
    -0.24
    pins
    -0.24
    POSITIVE LOGITS
    å¦Ĥä½ķ
    0.33
     future
    0.30
     how
    0.30
    -hide
    0.29
    æľªæĿ¥çļĦ
    0.29
    hei
    0.27
    Ñħи
    0.27
     sustain
    0.27
    geh
    0.26
    æľªæĿ¥
    0.26
    Activations Density 0.925%

    No Known Activations