© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 8302
    Prev
    Next
    INDEX
    Explanations

    to the point

    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
    éĢĨ
    -0.10
    æĿ¥èĩª
    -0.10
    轧
    -0.09
    ruk
    -0.09
    ureka
    -0.09
    Ñģке
    -0.09
    INCT
    -0.09
    IDA
    -0.09
     ÑĥвеÑĢен
    -0.09
    å½ĺ
    -0.08
    POSITIVE LOGITS
    éĤ£éĩĮ
    0.10
    æĮ¤
    0.09
    æ°´æºIJ
    0.09
    æľ«
    0.09
    è£ī
    0.09
    åIJİçļĦ
    0.09
    éĤ£åĦ¿
    0.09
     grips
    0.08
    æŃ£çīĪ
    0.08
    .saved
    0.08
    Activations Density 0.053%

    No Known Activations