© 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. 52213
    Prev
    Next
    INDEX
    Explanations

    quotation marks followed by capitalized words

    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
    a
    -0.20
    s
    -0.18
    à¸Ļ
    -0.17
    er
    -0.17
    y
    -0.15
    o
    -0.15
    es
    -0.14
    ãĥ£
    -0.14
    i
    -0.14
    en
    -0.14
    POSITIVE LOGITS
    imately
    0.15
    tempts
    0.13
    ings
    0.11
    INGS
    0.11
    bidden
    0.11
    lectual
    0.11
    tered
    0.10
    rition
    0.10
    ments
    0.10
    ذا
    0.10
    Activations Density 0.185%

    No Known Activations