© 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. 52777
    Prev
    Next
    INDEX
    Explanations

    wrong or right

    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
    é«ĺ级
    -0.10
    ãģ¦ãģĦ
    -0.09
    â¸
    -0.09
    éľ²
    -0.09
    à¸Ĥà¸Ļ
    -0.09
    å¯Ĵ
    -0.09
    grily
    -0.09
    ени
    -0.08
    enie
    -0.08
    æ¸ħåĩī
    -0.08
    POSITIVE LOGITS
     wrong
    0.14
    éĶĻäºĨ
    0.13
    没éĶĻ
    0.13
    wrong
    0.12
     right
    0.12
    ä¸į对
    0.12
     correct
    0.12
    æŃ£ç¡®
    0.12
     correctness
    0.11
    Wrong
    0.11
    Activations Density 0.059%

    No Known Activations