© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 50891
    Prev
    Next
    INDEX
    Explanations

    populate data structure

    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
    ernet
    -0.14
    æ»ĵ
    -0.09
    è·
    -0.08
    uffles
    -0.08
    æĺİçĻ½
    -0.08
    çĢijå¸ĥ
    -0.08
    ort
    -0.08
    è¿ŀ线
    -0.08
    atter
    -0.08
    av
    -0.08
    POSITIVE LOGITS
    家åĸ»æĪ·æĻĵ
    0.11
    ¨
    0.11
    ALLY
    0.10
    chod
    0.10
     Idol
    0.09
    ric
    0.09
    iciel
    0.09
    agation
    0.09
    迤
    0.09
    ìļĶ
    0.09
    Activations Density 0.072%

    No Known Activations