© 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. 58010
    Prev
    Next
    INDEX
    Explanations

    wrangler, wrapping, writhing, wringing, wrappers, wrangling

    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
    emand
    -0.10
    ohn
    -0.10
    ãĥ£
    -0.09
    arith
    -0.09
    oral
    -0.09
    stial
    -0.09
    ikki
    -0.09
    overe
    -0.09
    è²Į
    -0.09
    è¦ģçľĭ
    -0.09
    POSITIVE LOGITS
    icator
    0.10
    幺
    0.09
    abouts
    0.09
    ained
    0.09
    绶
    0.09
    èĶĵ
    0.09
    ض
    0.09
    cient
    0.08
    ç¼°
    0.08
    orio
    0.08
    Activations Density 0.013%

    No Known Activations