© 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. 31-TRANSCODER-HP
    7. 19445
    Prev
    Next
    INDEX
    Explanations

    the letter b

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     B
    -0.63
    çİĭ
    -0.58
    B
    -0.54
    _B
    -0.53
    	B
    -0.50
    *B
    -0.49
    ï¼¢
    -0.48
    >B
    -0.47
    ,B
    -0.46
    çİĭæŁIJ
    -0.46
    POSITIVE LOGITS
     b
    0.94
    b
    0.83
    *b
    0.64
    =b
    0.61
     б
    0.60
    ;b
    0.59
    :b
    0.58
    	b
    0.58
    ï½Ĥ
    0.57
    >b
    0.57
    Activations Density 12.599%

    No Known Activations