© 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. 4004
    Prev
    Next
    INDEX
    Explanations

    not a

    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
    ä¸įåĪ°
    -0.10
     Probably
    -0.09
    ç»Ģ
    -0.09
    ìĨį
    -0.09
     probably
    -0.09
     غÙĬر
    -0.08
    æĺ¯éĿŀ
    -0.08
    probably
    -0.08
    åĪ«
    -0.08
    ä¿Ĺ称
    -0.08
    POSITIVE LOGITS
    èĢĮæĺ¯
    0.13
    æľīä»»ä½ķ
    0.13
    epad
    0.11
    å·§åIJĪ
    0.11
     necessarily
    0.11
     anymore
    0.10
    游ä¹IJ
    0.10
    ä¹Łç®Ĺ
    0.09
     Chew
    0.09
     sondern
    0.09
    Activations Density 0.027%

    No Known Activations