© 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. 20738
    Prev
    Next
    INDEX
    Explanations

    usually not bad

    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
    6
    -0.09
    peak
    -0.09
    utex
    -0.08
    liner
    -0.08
    2
    -0.08
    3
    -0.08
    4
    -0.08
    对æłĩ
    -0.08
    è¿İ
    -0.08
    1
    -0.08
    POSITIVE LOGITS
    ä½łåĸľæ¬¢
    0.11
     полÑĮзÑĥ
    0.11
    带æĿ¥æĽ´å¤ļ
    0.11
    çļĦåį±å®³
    0.11
    å·¥ç¨ĭè´¨éĩı
    0.11
    严éĩįå½±åĵį
    0.11
    åģ¥åº·
    0.11
    è´ŁéĿ¢å½±åĵį
    0.10
    ä¸į太好
    0.10
    好å¤Ħ
    0.10
    Activations Density 0.263%

    No Known Activations