© 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. 53634
    Prev
    Next
    INDEX
    Explanations

    analysis of

    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
    çĦ¶æĺ¯
    -0.16
    ality
    -0.14
    sWith
    -0.12
    tat
    -0.11
    tings
    -0.11
    zilla
    -0.11
    erator
    -0.11
    arto
    -0.10
    ../../../
    -0.10
    ishly
    -0.10
    POSITIVE LOGITS
    sing
    0.11
    igator
    0.10
    cope
    0.09
    es
    0.09
    åijĺ
    0.09
    çłĶåĪ¤
    0.09
    аÑĤоÑĢ
    0.09
    人士
    0.09
    conda
    0.08
    /debug
    0.08
    Activations Density 0.068%

    No Known Activations