© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 37525
    Prev
    Next
    INDEX
    Explanations

    that being said

    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
    èĦį
    -0.09
    æĻ°
    -0.09
    åįħ
    -0.08
    å¹²èĦĨ
    -0.08
    ä¸įçŁ¥ä¸įè§ī
    -0.08
    ç®Ģ缴æĺ¯
    -0.08
    ç®ĢçĽ´å°±æĺ¯
    -0.08
    åľī
    -0.08
     figure
    -0.08
    ifold
    -0.08
    POSITIVE LOGITS
    è¿Ļä¹Ī说
    0.12
     nonetheless
    0.11
    ä¸į说
    0.10
    è¿Ļæł·è¯´
    0.10
    è¯Ŀ说
    0.10
     nevertheless
    0.09
    說
    0.09
     said
    0.09
    说
    0.09
    ä¾ĭå¤ĸ
    0.09
    Activations Density 0.034%

    No Known Activations