© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 3076
    Prev
    Next
    INDEX
    Explanations

    raid, destroyed, attack, fortress

    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
    åıįèħIJåĢ¡å»ī
    -0.10
    宫殿
    -0.10
    ç¼°
    -0.10
    ç¥ĿæĦ¿
    -0.09
     stadium
    -0.08
    ç¾ŀ
    -0.08
    殿
    -0.08
    åĪ»èĭ¦
    -0.08
    æĺı
    -0.08
    alendar
    -0.08
    POSITIVE LOGITS
    åı¯è§Ĥ
    0.11
    éªļæī°
    0.11
    æݳ
    0.11
    æİł
    0.09
    常æĢģåĮĸ
    0.09
    æĬ¥å¤į
    0.09
    忽
    0.09
    夷
    0.09
    è¿Ļ个è¡Įä¸ļ
    0.08
    çĶ·åŃ©åŃIJ
    0.08
    Activations Density 0.091%

    No Known Activations