© 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. 49837
    Prev
    Next
    INDEX
    Explanations

    unlikely or rare

    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
    æīįç®Ĺ
    -0.09
    borg
    -0.09
    åľ¨å®¶éĩĮ
    -0.09
     blind
    -0.08
    èĭ¦
    -0.08
    blind
    -0.08
    è£Ļ
    -0.08
    chet
    -0.08
    éŃĶ鬼
    -0.08
    ä¸įåĨįæĺ¯
    -0.08
    POSITIVE LOGITS
    unlikely
    0.22
     unlikely
    0.21
    åı¯èĥ½æĢ§
    0.14
    ä¸įåı¯èĥ½
    0.14
     likelihood
    0.13
     doubtful
    0.13
    æĺ¯ä¸įä¼ļ
    0.12
     impro
    0.12
    å¾Īæľīåı¯èĥ½
    0.11
     improbable
    0.11
    Activations Density 0.048%

    No Known Activations