© 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. 37390
    Prev
    Next
    INDEX
    Explanations

    Vanderbilt University

    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
    esthetic
    -0.10
    ongan
    -0.09
    lining
    -0.09
    æ¾Ī
    -0.09
    ber
    -0.09
    ×ķת
    -0.09
    yun
    -0.09
    oader
    -0.08
    est
    -0.08
    isten
    -0.08
    POSITIVE LOGITS
    ÙĤ
    0.10
    åŃIJéĩĮ
    0.10
    standing
    0.10
    unts
    0.08
    çĩ¥
    0.08
    oger
    0.08
    亲çĪ±
    0.08
    æĩ¦
    0.08
    çķ¸å½¢
    0.08
    ç²Ĵ
    0.08
    Activations Density 0.002%

    No Known Activations