© 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. 9948
    Prev
    Next
    INDEX
    Explanations

    to outline, to measure, to recognise, to edit

    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
    atee
    -0.08
    åħ¶ä¸Ń
    -0.08
    å¡Ķ
    -0.08
    yms
    -0.08
    åĩłçĤ¹
    -0.08
    bens
    -0.08
    øre
    -0.08
    ï¼ļ</
    -0.08
    _tac
    -0.08
    ä¸īçĤ¹
    -0.08
    POSITIVE LOGITS
     nämlich
    0.11
    好çľĭçļĦ
    0.10
    建设åĴĮ
    0.10
    è´¨éĩıåĴĮ
    0.09
    çľ¼åħī
    0.09
     namely
    0.09
    æľīå¤ļ大
    0.09
    å¾Ĺèµ·
    0.09
    éĤ£å°±æĺ¯
    0.09
    éĹ´çļĦ
    0.08
    Activations Density 0.136%

    No Known Activations