© 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. 22343
    Prev
    Next
    INDEX
    Explanations

    change and replace

    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
    åľ¨æİ¥åıĹ
    -0.10
    ä¸ĢåĽ¢
    -0.09
    为èĩªå·±
    -0.09
    çļĦ家åºŃ
    -0.09
    _VO
    -0.09
    avel
    -0.08
    argout
    -0.08
    åįķä»·
    -0.08
    å°½æĥħ
    -0.08
    对èĩªå·±
    -0.08
    POSITIVE LOGITS
    åıĬæŶ
    0.12
    ç»´æĬ¤
    0.11
    æ¯ıéļĶ
    0.11
    éĢĤæŶ
    0.10
    åĭ¤
    0.10
    æĽ´æį¢
    0.10
    cek
    0.10
    å¹²é¢Ħ
    0.10
    æ£ĢæŁ¥
    0.10
    lime
    0.10
    Activations Density 0.115%

    No Known Activations