© 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. 33-TRANSCODER-HP
    7. 54267
    Prev
    Next
    INDEX
    Explanations

    this

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     defaults
    -0.27
     repeats
    -0.26
     is
    -0.25
     ends
    -0.25
    Containing
    -0.25
    äºķ
    -0.25
     fixes
    -0.24
     redirects
    -0.24
    isks
    -0.24
     results
    -0.24
    POSITIVE LOGITS
     Most
    0.36
     type
    0.36
     ever
    0.35
     most
    0.35
    most
    0.33
    ç±»åŀĭçļĦ
    0.31
    type
    0.31
     decreasing
    0.30
    $type
    0.30
    	type
    0.29
    Activations Density 0.027%

    No Known Activations