© 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. 22085
    Prev
    Next
    INDEX
    Explanations

    getting stuck

    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
    让å®ĥ
    -0.12
    åĬłä»¥
    -0.10
    eos
    -0.10
    æĪijåıĪ
    -0.10
    æĬĬå®ĥ
    -0.09
    åĨĢ
    -0.09
    ijľ
    -0.09
    使ä¹ĭ
    -0.09
    chooser
    -0.09
    allon
    -0.09
    POSITIVE LOGITS
    æł¹æľ¬æ²¡æľī
    0.09
     excess
    0.08
     indeed
    0.08
     FAC
    0.08
    ä¹ĭåĬ¿
    0.08
    âĢŀ
    0.07
    æł¹æľ¬
    0.07
    çļĦå°±æĺ¯
    0.07
     przec
    0.07
    ron
    0.07
    Activations Density 0.107%

    No Known Activations