© 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. 30473
    Prev
    Next
    INDEX
    Explanations

    reduce negative impact

    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
    hyth
    -0.11
    ä¸įæŃ¢
    -0.09
    å¹ķ
    -0.09
     Bowman
    -0.09
    å½Ĵ
    -0.08
    æĺ¯æĢİä¹Ī
    -0.08
     Clair
    -0.08
     Breath
    -0.08
    á»ĭ
    -0.08
    åįĵ
    -0.08
    POSITIVE LOGITS
    /mit
    0.12
    éĻįä½İæĪIJæľ¬
    0.10
    è±ķ
    0.09
    /MIT
    0.09
    ä¸įå¿ħè¦ģçļĦ
    0.09
     chance
    0.09
     Chance
    0.09
     garant
    0.09
     guarantees
    0.08
    /min
    0.08
    Activations Density 0.059%

    No Known Activations