© 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. 18085
    Prev
    Next
    INDEX
    Explanations

    more than, higher than, as many as, much higher than, far exceeding, cheaper than, lower than, significantly higher than, twice that 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
    å§ĭç»Īä¿ĿæĮģ
    -0.10
    被æĬĵ
    -0.09
     nisi
    -0.09
    就好åĥı
    -0.09
    าว
    -0.08
    OTH
    -0.08
    éĥĦ
    -0.08
    èĢģå©Ĩ
    -0.08
    礼æĭľ
    -0.08
    çĹ¼
    -0.08
    POSITIVE LOGITS
    é¢ij
    0.09
    ghi
    0.09
    大å¦Ī
    0.09
    è¿ľ
    0.08
    ä¸įåIJĮäºİ
    0.08
     обÑıзан
    0.08
    å®ŀåľ¨
    0.08
    æīĭ游
    0.08
    utos
    0.08
    éģ¥
    0.08
    Activations Density 0.112%

    No Known Activations