© 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. 1134
    Prev
    Next
    INDEX
    Explanations

    preliminary studies and findings

    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
    æĮģç»Ńæİ¨è¿Ľ
    -0.09
    è´Ł
    -0.09
    ç»Ħç»ĩå®ŀæĸ½
    -0.09
     Sawyer
    -0.09
    论述
    -0.08
    incess
    -0.08
    è®¤çľŁèIJ½å®ŀ
    -0.08
    åĿļæĮģä¸įæĩĪ
    -0.08
    éĺ»
    -0.08
    éºĪ
    -0.08
    POSITIVE LOGITS
    åĪĿæŃ¥
    0.22
     preliminary
    0.13
    æµ®
    0.11
    ç²Ĺ
    0.11
     Prel
    0.11
    éĺ¶æ®µ
    0.10
     prelim
    0.10
    ä¹ĭåĪĿ
    0.09
     glance
    0.09
     disen
    0.09
    Activations Density 0.069%

    No Known Activations