© 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. 10851
    Prev
    Next
    INDEX
    Explanations

    eyes, pupil, wide, unilateral

    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
    åŀĴ
    -0.09
     gar
    -0.08
     Eb
    -0.08
     Gib
    -0.08
     Err
    -0.08
    endar
    -0.07
     jun
    -0.07
    æĸ¥
    -0.07
    åij£
    -0.07
    gar
    -0.07
    POSITIVE LOGITS
    çļĦçľ¼çĿĽ
    0.25
    çľ¼çĿĽ
    0.23
     eyes
    0.19
    åıĮçľ¼
    0.18
    çŀ³
    0.17
     глаз
    0.16
    eyes
    0.16
    çľ¼çIJĥ
    0.16
    çľ¼
    0.16
    羸
    0.16
    Activations Density 0.061%

    No Known Activations