© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 8698
    Prev
    Next
    INDEX
    Explanations

    end-image processing

    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
    ören
    -0.09
    ern
    -0.08
    cko
    -0.08
    .mx
    -0.08
    ansion
    -0.08
    inge
    -0.07
    Kick
    -0.07
    django
    -0.07
    rid
    -0.07
    ños
    -0.07
    POSITIVE LOGITS
    åħ³äºİæĪij们
    0.12
    SystemService
    0.10
    oodles
    0.10
    åĽ½éĻħè´¸æĺĵ
    0.09
    guard
    0.09
     pars
    0.09
    æĸ°è¥¿
    0.09
    å¸ĤåĮº
    0.09
    yard
    0.09
    æĿ¥ä¸įåıĬ
    0.09
    Activations Density 0.587%

    No Known Activations