© 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. 52151
    Prev
    Next
    INDEX
    Explanations

    physics of

    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
    泫
    -0.10
    è®®äºĭ
    -0.10
    æĦıæĢĿæĺ¯
    -0.10
    æijŃ
    -0.09
     scheme
    -0.09
    æ»ĵ
    -0.09
    çļĦç¼ĺæķħ
    -0.09
     decks
    -0.09
    åĵĴ
    -0.09
    cheme
    -0.09
    POSITIVE LOGITS
    bard
    0.09
    å¢ŀéĩı
    0.09
     ÑģлÑĥÑĩаÑı
    0.08
    ела
    0.08
    æī¶è´«å·¥ä½ľ
    0.08
    ç²¾éĢļ
    0.08
     consent
    0.08
    .addHandler
    0.08
    bler
    0.08
    ัà¸Ķ
    0.08
    Activations Density 0.223%

    No Known Activations