© 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. 23064
    Prev
    Next
    INDEX
    Explanations

    act or serve as

    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
    egot
    -0.09
    æĤ¯
    -0.08
    é©¿
    -0.08
    è¡¡
    -0.08
    éĻĢ
    -0.08
    _centers
    -0.08
    нÑĮ
    -0.08
    谢谢
    -0.08
    å°Ĭ严
    -0.08
     extract
    -0.08
    POSITIVE LOGITS
    ä½ľä¸º
    0.19
    ä½ľçĤº
    0.16
     sebagai
    0.14
    åħħå½ĵ
    0.14
    ä½ľä¸ºä¸Ģ个
    0.13
     каÑĩеÑģÑĤве
    0.13
     jako
    0.13
    å½ĵåģļ
    0.12
     als
    0.12
    isan
    0.12
    Activations Density 0.095%

    No Known Activations