© 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. 34999
    Prev
    Next
    INDEX
    Explanations

    excerpt from

    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
    讳
    -0.10
    ning
    -0.08
    6
    -0.08
    headed
    -0.08
    /or
    -0.08
    ateral
    -0.08
    çķĮçļĦ
    -0.08
     separate
    -0.08
    ampo
    -0.08
    utton
    -0.08
    POSITIVE LOGITS
    æijĺ
    0.11
    ä¸Ģäºĭ
    0.10
    å¦Ĭ
    0.09
    виÑĤ
    0.09
    æijĺç¼ĸ
    0.09
    IFF
    0.09
    夫
    0.09
     ÑģÑĤан
    0.09
    æĬĵ好
    0.09
    西äºļ
    0.09
    Activations Density 0.007%

    No Known Activations