© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 53158
    Prev
    Next
    INDEX
    Explanations

    numbers followed by parts, stages, or acts

    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
    岫
    -0.09
    .va
    -0.09
    ç¨
    -0.09
    ç¼Ń
    -0.09
    incerely
    -0.09
    ously
    -0.08
    _Dep
    -0.08
    åĬ¾
    -0.08
    ylko
    -0.08
    quate
    -0.08
    POSITIVE LOGITS
     parts
    0.13
    -parts
    0.11
     ÑĩаÑģÑĤи
    0.10
    parts
    0.10
     Parts
    0.10
    éĴ¤
    0.09
    ISTS
    0.09
    _parts
    0.09
     pieces
    0.09
    大èĤ¡ä¸ľ
    0.08
    Activations Density 0.076%

    No Known Activations