© 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. 51657
    Prev
    Next
    INDEX
    Explanations

    modal verbs like may, can, have

    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
    裸
    -0.09
    erea
    -0.09
    éŃĶçİĭ
    -0.09
    æijĦ
    -0.09
    çªģå¦Ĥåħ¶æĿ¥
    -0.09
    åı¯ä¸įæĺ¯
    -0.08
    èĻ½çĦ¶
    -0.08
    ãģĮãĤī
    -0.08
    ä¸Ģç¯ĩ
    -0.08
    oken
    -0.08
    POSITIVE LOGITS
    为ä¹ĭ
    0.10
    oda
    0.10
    odge
    0.09
    ä¸IJ
    0.09
    ç§į
    0.08
    åħ¨ç¨ĭ
    0.08
    å¾Īå¤ļäºĭæĥħ
    0.08
     даж
    0.08
    çļ´
    0.08
    次
    0.08
    Activations Density 0.342%

    No Known Activations