© 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. 50457
    Prev
    Next
    INDEX
    Explanations

    following along with

    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
    -offs
    -0.09
    åĩºçĤī
    -0.09
    éĻIJéĩı
    -0.09
    åŁ´
    -0.09
    izons
    -0.09
    çĤī
    -0.09
    çļ¿
    -0.09
    们
    -0.08
     PageInfo
    -0.08
    çĦ¶æĺ¯
    -0.08
    POSITIVE LOGITS
    .RunWith
    0.10
     ÑģебÑı
    0.10
    èĬĤ缮ä¸Ń
    0.10
    ewis
    0.09
    ue
    0.09
    eur
    0.09
    __$
    0.09
    ueur
    0.09
    ood
    0.09
    ippi
    0.09
    Activations Density 0.016%

    No Known Activations