© 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. 22797
    Prev
    Next
    INDEX
    Explanations

    cracking or cracks

    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
     pháºŃn
    -0.10
    aka
    -0.09
     tact
    -0.08
    泡
    -0.08
    UnderTest
    -0.08
    åѦä½į
    -0.08
    ouv
    -0.08
    /man
    -0.08
    atic
    -0.08
    èĪĨ
    -0.08
    POSITIVE LOGITS
    pot
    0.10
    blink
    0.10
    ä¹Ŀé¾Ļ
    0.09
    ERY
    0.09
    usi
    0.09
    丼
    0.09
    å¼ĢæĿ¥
    0.09
    ru
    0.09
    马路
    0.09
    tober
    0.09
    Activations Density 0.014%

    No Known Activations