© 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. 19608
    Prev
    Next
    INDEX
    Explanations

    Bo followed by 'okeh', 'oming', 'ra', 'ing'

    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
    cion
    -0.09
    ues
    -0.09
    ollar
    -0.09
    æ³ķå®ļ代表
    -0.09
    pire
    -0.08
    æĽ´å¤ļä¿¡æģ¯
    -0.08
    ophon
    -0.08
    ãģĵãģ¨
    -0.08
    å¤ĦåĪĨ
    -0.08
    gram
    -0.08
    POSITIVE LOGITS
    iful
    0.11
    ï¸
    0.10
    imet
    0.10
    enheim
    0.09
    沿线
    0.09
    strap
    0.09
    est
    0.09
    LOPT
    0.09
    å¨ĺ
    0.09
    iled
    0.09
    Activations Density 0.027%

    No Known Activations