© 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. 36466
    Prev
    Next
    INDEX
    Explanations

    Berkeley

    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
    ä¸ĭåİ»
    -0.10
    iele
    -0.10
    agra
    -0.09
    åŃIJéĩĮ
    -0.09
    sWith
    -0.09
    æ½¢
    -0.09
    åij¼åºĶ
    -0.09
     lÆ°ng
    -0.09
    stell
    -0.09
    ìľ¨
    -0.08
    POSITIVE LOGITS
    åºľ
    0.10
    à´´
    0.10
     Irvine
    0.09
    èħĭ
    0.08
    eres
    0.08
     Heights
    0.08
    efs
    0.08
    erase
    0.08
    voir
    0.08
    avr
    0.08
    Activations Density 0.006%

    No Known Activations