© 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. 54298
    Prev
    Next
    INDEX
    Explanations

    credits and attribution

    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
    æĬķèµĦé¢Ŀ
    -0.10
     pres
    -0.09
    åĨľä¸ļåĨľæĿij
    -0.09
    нÑĤеÑĢ
    -0.09
    arget
    -0.08
    ä»İ容
    -0.08
    wick
    -0.08
    æĤłéĹ²
    -0.08
    举æŃ¢
    -0.08
    è§Ħ
    -0.08
    POSITIVE LOGITS
    ç½²
    0.17
     Credits
    0.14
     credits
    0.14
    _credit
    0.13
    .credit
    0.13
    鸣
    0.13
    Credits
    0.13
     attribution
    0.12
    credits
    0.12
    credit
    0.11
    Activations Density 0.163%

    No Known Activations