© 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. 23940
    Prev
    Next
    INDEX
    Explanations

    giving an answer

    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
    ë°
    -0.09
    tle
    -0.09
    宽æĿ¾
    -0.09
    ếp
    -0.08
    ently
    -0.08
    pong
    -0.08
    oub
    -0.08
    amba
    -0.08
    .reload
    -0.08
    abouts
    -0.08
    POSITIVE LOGITS
    è¿«ä¸įåıĬå¾ħ
    0.11
    strup
    0.10
    æݨéĢģ
    0.09
    è±¹
    0.09
    å§Ķç»Ħç»ĩ
    0.08
    fel
    0.08
    åĪĨ为
    0.08
    âľį
    0.08
    æĦıæĦ¿
    0.08
    _TYP
    0.08
    Activations Density 0.028%

    No Known Activations