© 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. 62123
    Prev
    Next
    INDEX
    Explanations

    cr followed by 'annies', 'anked', 'isco', 'imp', 'ink', 'è', 'aters', 'amps'

    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
    ä¼Ļ
    -0.10
    ä¸Ģåı£
    -0.09
    asca
    -0.09
    owy
    -0.09
    zed
    -0.09
    7
    -0.08
    0
    -0.08
    èħĮ
    -0.08
    9
    -0.08
    8
    -0.08
    POSITIVE LOGITS
    åľ°å¸¦
    0.11
    æŁ¬
    0.11
    æ¹¾
    0.10
    象
    0.09
    lac
    0.09
    æĦŁæ¿Ģ
    0.09
    æĪ¢
    0.09
    éĶ¡
    0.09
    lace
    0.09
    ultimate
    0.08
    Activations Density 0.011%

    No Known Activations