© 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. 61200
    Prev
    Next
    INDEX
    Explanations

    ambush attacks

    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
    仲
    -0.09
    ÙĬص
    -0.09
    (ok
    -0.09
    ernes
    -0.08
    atie
    -0.08
    çĻ«
    -0.08
    å¥Ķæ³¢
    -0.08
     Fla
    -0.08
    èĩªçIJĨ
    -0.08
    æĿĥå¨ģ
    -0.08
    POSITIVE LOGITS
    个åĪ«
    0.10
    èĥĮåIJİ
    0.10
    ä¸įç¡®å®ļæĢ§
    0.09
    èĤĭ
    0.09
    ä¸ĭæīĭ
    0.09
    ç®Ń
    0.09
    ntax
    0.09
    æIJ¦
    0.09
    èĸ®
    0.09
     ambush
    0.09
    Activations Density 0.061%

    No Known Activations