© 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. 9836
    Prev
    Next
    INDEX
    Explanations

    good enough

    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
    æ¶Ĥ
    -0.10
    æĪ´
    -0.10
    æijĦ
    -0.10
    æĨ¬
    -0.09
    å·½
    -0.09
     REM
    -0.09
    allen
    -0.09
    PACK
    -0.09
    scal
    -0.08
    åįĥåı¤
    -0.08
    POSITIVE LOGITS
    ä¸įå¤Ł
    0.14
     acceptable
    0.12
    å¤Ł
    0.11
    è¿ĺä¸įå¤Ł
    0.11
    å¤ł
    0.11
    acceptable
    0.10
    å«Įå¼ĥ
    0.10
    è¶³å¤Ł
    0.09
     mức
    0.09
    满æĦı
    0.09
    Activations Density 0.055%

    No Known Activations