© 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. 53742
    Prev
    Next
    INDEX
    Explanations

    victory

    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
    å¹³çŃī
    -0.12
    ableObject
    -0.09
    è·Ł
    -0.09
     naw
    -0.09
    åIJĮçŃī
    -0.09
    ç´§è·Ł
    -0.09
    注æĺİåĩºå¤Ħ
    -0.08
    éĢĢ
    -0.08
     balancing
    -0.08
    èľ¡
    -0.08
    POSITIVE LOGITS
     margins
    0.11
    -margin
    0.11
    ä¼ĺè¶Ĭ
    0.10
    æĹ©æĹ©
    0.10
    å·®è·Ŀ
    0.10
    è¹Ĥ
    0.10
     EAR
    0.10
    steam
    0.09
    çļĦä¼ĺåĬ¿
    0.09
    æij§
    0.09
    Activations Density 0.252%

    No Known Activations