© 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. 62391
    Prev
    Next
    INDEX
    Explanations

    or either

    np_acts-logits-general · gemini-2.5-flash-lite

    Medical and health-related content — especially disease names, symptoms, treatments, and preventive-care information.

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31
    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
     annum
    -0.10
    æĭīåĬ¨
    -0.08
    å®Įæ¯ķ
    -0.08
    头çĸ¼
    -0.08
    æ¯ıæ¯ı
    -0.08
    ãģªãĤĵãģĭ
    -0.08
    çģµåĬ¨
    -0.07
    olean
    -0.07
    ä¸įåı¯æĢĿè®®
    -0.07
    ä¹ĥ
    -0.07
    POSITIVE LOGITS
    å°ıä¼Ļä¼´
    0.09
    å¸®ä½ł
    0.09
    æĽ¿è¡¥
    0.08
     kinky
    0.08
    çĪ±ä½ł
    0.08
     ÐłÐ¾
    0.08
     nons
    0.08
    èĽ²
    0.08
    enso
    0.08
    ä½łå°±
    0.08
    Activations Density 0.241%

    No Known Activations