© 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. 39587
    Prev
    Next
    INDEX
    Explanations

    part number or roman numeral

    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
    åģıåģı
    -0.10
    ered
    -0.09
    elen
    -0.08
    å¥Ī
    -0.08
    LOY
    -0.08
    碰
    -0.08
    æĻ®éģį
    -0.08
    碰åĪ°
    -0.08
    olas
    -0.08
     SYMBOL
    -0.08
    POSITIVE LOGITS
     Part
    0.10
     part
    0.10
     Ча
    0.10
    _jet
    0.09
    Part
    0.09
    åħĪ导
    0.09
    tte
    0.09
    part
    0.09
    ä¸ĸçķĮè§Ĥ
    0.09
    terior
    0.09
    Activations Density 0.045%

    No Known Activations