© 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. 24145
    Prev
    Next
    INDEX
    Explanations

    forbidden or allowed

    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
    Negative Logits
    çĻ½éĽª
    -0.10
    æĬ¢
    -0.09
    ä½ĵä¼ļåĪ°
    -0.09
    è¶ģ
    -0.09
     Ever
    -0.09
    æħij
    -0.08
    éļ½
    -0.08
    æ·±åĪĩ
    -0.08
     *)__
    -0.08
    Ever
    -0.08
    POSITIVE LOGITS
    ä¸įåħģ许
    0.20
    ç¦ģæŃ¢
    0.19
    åħģ许
    0.16
     forbidden
    0.16
    严ç¦ģ
    0.15
     allowed
    0.15
    allowed
    0.14
     Allowed
    0.13
    Allowed
    0.13
    _allowed
    0.13
    Activations Density 0.158%

    No Known Activations