© 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. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 1473
    Prev
    Next
    INDEX
    Explanations

    YT followed by symbol/letter

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-32b-saes/saes_Qwen_Qwen3-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    峪
    -0.11
    à§į
    -0.11
    chest
    -0.11
    ÑĦи
    -0.10
    èĢĺ
    -0.10
    _yaml
    -0.10
    æľīæĦıä¹ī
    -0.09
    hip
    -0.09
    æĪij以为
    -0.09
    .au
    -0.09
    POSITIVE LOGITS
    ields
    0.16
    osemite
    0.13
    IELD
    0.11
    ron
    0.11
    andex
    0.11
    lish
    0.10
    год
    0.10
    annis
    0.10
    esterday
    0.10
    amaha
    0.10
    Activations Density 0.840%

    No Known Activations