© 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. 21891
    Prev
    Next
    INDEX
    Explanations

    subtract, traction, distraction

    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
    èĶ¼
    -0.10
    åĸĩåıŃ
    -0.09
     Loud
    -0.09
    ÙĬÙĪÙĨ
    -0.09
    ration
    -0.09
    éĢĢåĩº
    -0.09
    ulators
    -0.08
    iment
    -0.08
    draft
    -0.08
    acas
    -0.08
    POSITIVE LOGITS
    ä¸Ģç§Ĵ
    0.09
    害
    0.09
    æĢ»ä½ĵ
    0.09
    çļĦæķ´ä½ĵ
    0.09
    åİŁæľīçļĦ
    0.09
     overall
    0.08
    夺
    0.08
     detr
    0.08
    ends
    0.08
    OCR
    0.08
    Activations Density 0.007%

    No Known Activations