© 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. 51433
    Prev
    Next
    INDEX
    Explanations

    rate of

    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
    äºĨä¸Ģ份
    -0.11
    å½ĸ
    -0.10
    åĵ¨
    -0.09
    ä¹ħäºĨ
    -0.09
    imed
    -0.09
    åĮ¿
    -0.08
    åĽ½å®¶å®īåħ¨
    -0.08
    .ResponseBody
    -0.08
    äºĨä¸Ģæī¹
    -0.08
    Warehouse
    -0.08
    POSITIVE LOGITS
    lander
    0.10
     rates
    0.10
     Rates
    0.09
    èĬĤå¥ı
    0.09
    -paced
    0.09
     pace
    0.09
     rate
    0.09
     Rate
    0.09
     Pace
    0.09
    (rate
    0.09
    Activations Density 0.074%

    No Known Activations