© 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. 50782
    Prev
    Next
    INDEX
    Explanations

    evergreen

    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
    æŃ£åľ¨è¿Ľè¡Į
    -0.09
    èĽĭçĻ½
    -0.09
    etre
    -0.08
    è¡Ģ管
    -0.08
    anela
    -0.08
    çĨ¬
    -0.08
    ä¼
    -0.08
    çģ«çĥŃ
    -0.08
    iotic
    -0.08
    otion
    -0.08
    POSITIVE LOGITS
    åĨĴåĩº
    0.10
     Humanity
    0.09
     sty
    0.09
    ä¹ĭ称
    0.09
    èĸħ
    0.08
    èķ¤
    0.08
    autos
    0.08
    lude
    0.08
    entially
    0.08
    (runtime
    0.08
    Activations Density 0.006%

    No Known Activations