© 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. 55336
    Prev
    Next
    INDEX
    Explanations

    regex pattern extraction

    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
    æĹłèģĬ
    -0.11
    涩
    -0.10
     Flush
    -0.10
    æµĩ
    -0.09
    IfExists
    -0.09
    _strerror
    -0.09
    件
    -0.09
     Bundy
    -0.09
    ç¿Ĭ
    -0.09
    à¸ķำ
    -0.09
    POSITIVE LOGITS
    .Pattern
    0.13
     pattern
    0.13
    (pattern
    0.12
    .Matcher
    0.12
     patterns
    0.12
     Pattern
    0.11
    Pattern
    0.11
    pattern
    0.11
    ãĥijãĤ¿ãĥ¼ãĥ³
    0.10
    attern
    0.10
    Activations Density 0.312%

    No Known Activations