© 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. 1528
    Prev
    Next
    INDEX
    Explanations

    watching or seeing

    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
    åijĬ
    -0.11
    Ùģار
    -0.10
    /question
    -0.10
    å¸ĸ
    -0.09
    otto
    -0.09
     Hind
    -0.09
    ãĥªãĤ¢
    -0.09
     rak
    -0.09
    uren
    -0.09
    åĵģç§į
    -0.09
    POSITIVE LOGITS
    åºĶä»ĺ
    0.10
    é£Ĵ
    0.09
    mani
    0.08
    å¥¹åľ¨
    0.08
    åŃIJéĩĮ
    0.08
    본
    0.08
    bare
    0.08
    â½Ģ
    0.08
     него
    0.08
    able
    0.08
    Activations Density 0.025%

    No Known Activations