© 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. 20568
    Prev
    Next
    INDEX
    Explanations

    winning streak or losing streak

    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
    æĺ¾
    -0.09
    ler
    -0.09
    ìĦ
    -0.08
     Crossing
    -0.08
     Stage
    -0.08
    IVO
    -0.08
    .yahoo
    -0.08
    é²ľæĺİ
    -0.08
    (zip
    -0.08
    åħ¨é¢Ŀ
    -0.08
    POSITIVE LOGITS
     series
    0.14
    éĵ¾æĿ¡
    0.12
    ãĤ·ãĥªãĥ¼ãĤº
    0.12
    series
    0.12
    ä¹ħ
    0.11
    ç³»åĪĹ
    0.11
     streak
    0.11
    è¿ŀèĥľ
    0.11
    çĭ¬è§Ĵ
    0.10
    åºıåĪĹ
    0.10
    Activations Density 0.076%

    No Known Activations