© 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. 5305
    Prev
    Next
    INDEX
    Explanations

    strange results, odd question, bizarre behavior

    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
    åŀĭ
    -0.10
    apl
    -0.09
     ÑģобÑģÑĤвенно
    -0.09
     Nunes
    -0.09
    owing
    -0.09
    andle
    -0.09
    asca
    -0.09
    æľįåĬ¡å¹³åı°
    -0.09
    èĦĤ
    -0.09
    uming
    -0.08
    POSITIVE LOGITS
    å¥ĩæĢª
    0.14
    åı¤æĢª
    0.13
    æĢª
    0.13
    å¥ĩå¼Ĥ
    0.11
     strange
    0.11
     odd
    0.10
    å¥ĩçī¹
    0.10
    å¥ĩå¦Ļ
    0.10
    è¹Ĭ
    0.10
    ãģĬãģĭ
    0.10
    Activations Density 0.080%

    No Known Activations