© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 19574
    Prev
    Next
    INDEX
    Explanations

    somatic

    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
    常å§Ķ
    -0.11
    /cop
    -0.10
    Ñĥл
    -0.10
    uard
    -0.09
    errupt
    -0.09
    磨
    -0.09
    ../
    -0.09
    ucc
    -0.09
    urs
    -0.09
    smith
    -0.09
    POSITIVE LOGITS
    çĦ¶æĺ¯
    0.12
    brero
    0.11
    mercial
    0.11
    ewhere
    0.11
    plete
    0.09
    è¡·
    0.09
    ars
    0.09
    æĹ¥åŃIJ
    0.09
    reon
    0.09
    erville
    0.09
    Activations Density 0.048%

    No Known Activations