© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 50980
    Prev
    Next
    INDEX
    Explanations

    Nixon or chloride or nginx

    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
    s
    -0.10
    sword
    -0.10
    ing
    -0.09
    -0.09
    sheets
    -0.09
    æķijæĬ¤
    -0.08
    åīįæıIJ
    -0.08
    ingly
    -0.08
    ungi
    -0.08
     QQ
    -0.08
    POSITIVE LOGITS
    ames
    0.11
    pile
    0.10
    佬
    0.10
    à¸Ńาศ
    0.10
    èĽĻ
    0.09
    oka
    0.09
    IAL
    0.09
    esting
    0.09
    ITE
    0.09
    gram
    0.09
    Activations Density 0.006%

    No Known Activations