© 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. 50071
    Prev
    Next
    INDEX
    Explanations

    numbers before 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
    ario
    -0.10
    ice
    -0.10
     Schro
    -0.10
    ãģ¤ãģ®
    -0.09
    æ¸
    -0.09
    骢
    -0.09
    ums
    -0.09
    vens
    -0.09
    orio
    -0.09
    egis
    -0.09
    POSITIVE LOGITS
    etwork
    0.09
    mart
    0.08
    maker
    0.08
    mars
    0.08
    ³
    0.08
    ished
    0.08
    erval
    0.08
    orst
    0.08
    º
    0.08
    emonic
    0.08
    Activations Density 0.291%

    No Known Activations