© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 20870
    Prev
    Next
    INDEX
    Explanations

    seriously, but, actually, literally

    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
     Rubin
    -0.10
    SCO
    -0.09
    emi
    -0.09
     Dodd
    -0.08
    ä¸įæĸŃåľ°
    -0.08
    ви
    -0.08
    áŀĸ
    -0.08
    omanip
    -0.08
     Sox
    -0.08
     Dere
    -0.08
    POSITIVE LOGITS
    Seriously
    0.12
    çİ°å®ŀä¸Ń
    0.10
     Seriously
    0.10
    ä¸įå¼Ģ
    0.10
    忸
    0.10
    è¿ĺæĺ¯è¦ģ
    0.09
     seriously
    0.09
    ãģ¾ãģĹãģ¦
    0.09
    è¿Ļä¹Ī说
    0.09
    åºļ
    0.09
    Activations Density 0.076%

    No Known Activations