© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 5002
    Prev
    Next
    INDEX
    Explanations

    crucial times or intervals

    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
    累积
    -0.11
    åı¯æĥ³
    -0.11
    湫
    -0.09
    éric
    -0.09
    æŃ©ãģĦãģ¦
    -0.08
    UNG
    -0.08
    ä¸ĢåŃĹ
    -0.08
    \Php
    -0.08
    å±ķå¼Ģ
    -0.08
    accumulate
    -0.08
    POSITIVE LOGITS
    æģ°å½ĵ
    0.16
    éĢĤæŶ
    0.15
    æĹ¶æľº
    0.15
    çļĦåħ³éĶ®
    0.13
    åħ³éĶ®æŶåĪ»
    0.12
    æģ°
    0.12
    ä»Ģä¹ĪæŶåĢĻ
    0.12
    åħ³éĶ®
    0.12
    çļĦæŶåĢĻ
    0.12
    åIJĪéĢĤçļĦ
    0.11
    Activations Density 0.064%

    No Known Activations