© 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. 57628
    Prev
    Next
    INDEX
    Explanations

    introducing a contrast or caveat

    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
    ê²½
    -0.09
    adera
    -0.09
    èIJį
    -0.09
    æĭ¦
    -0.09
    assage
    -0.09
    æīĵéĢļ
    -0.09
    CLEAR
    -0.08
    ×ķ×ij
    -0.08
     bánh
    -0.08
    èĢĮåIJİ
    -0.08
    POSITIVE LOGITS
    åıįæŃ£
    0.10
    è¿ĺæĺ¯æ¯Ķè¾ĥ
    0.09
    zin
    0.09
    å½ĵæŶ
    0.09
    个人è§ĤçĤ¹
    0.08
    _deinit
    0.08
     <![
    0.08
    æĭĴä¸į
    0.08
    éĢĨ
    0.08
    _exports
    0.08
    Activations Density 0.050%

    No Known Activations