© 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. 2774
    Prev
    Next
    INDEX
    Explanations

    cheat codes and maximum cost

    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
    á¹ĩ
    -0.10
    ä¸įæ¸ħ
    -0.09
    顶
    -0.09
    rapped
    -0.09
    踹
    -0.09
    éĿŀ常好
    -0.08
     Fantasy
    -0.08
    itur
    -0.08
    天ä¸Ĭ
    -0.08
    eren
    -0.08
    POSITIVE LOGITS
    è£Ļ
    0.09
    ahkan
    0.09
    ä½ľä¸ļ
    0.09
    æ·Ĩ
    0.09
    обÑī
    0.09
    ç§ijæķĻ
    0.09
    çĭ®
    0.09
    ackers
    0.09
    icles
    0.09
    zer
    0.08
    Activations Density 0.009%

    No Known Activations