© 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. 36688
    Prev
    Next
    INDEX
    Explanations

    never done

    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
    ä¸įåı¯
    -0.10
     blows
    -0.09
    ød
    -0.09
     Pier
    -0.08
    åIJ¹
    -0.08
    åı¯ä»¥ä½¿
    -0.08
    ä¸įåħį
    -0.08
    ä¸įå®ľ
    -0.08
     LAW
    -0.08
    yme
    -0.08
    POSITIVE LOGITS
    éĤ£ä¹Īå¤ļ
    0.13
    è¿Ļä¹Īå¤ļ
    0.13
    å¦ĤæѤ
    0.12
    éĤ£ä¹Ī容æĺĵ
    0.11
    æľīä»»ä½ķ
    0.11
    è¿Ļä¸ĢåĪĩ
    0.11
     вообÑīе
    0.10
     überhaupt
    0.10
    管å±Ģ
    0.10
    ä¿ĥæĪIJ
    0.10
    Activations Density 0.165%

    No Known Activations