© 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. 2057
    Prev
    Next
    INDEX
    Explanations

    know how / that / exactly

    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
    .currentTarget
    -0.10
    解说
    -0.09
    ä¸Ģä¸Ģ
    -0.09
    ilight
    -0.09
     Mor
    -0.09
    (AP
    -0.09
    ç»ĦåĽ¾
    -0.08
    大桥
    -0.08
    ç¬ķ
    -0.08
    arro
    -0.08
    POSITIVE LOGITS
    æ·±çŁ¥
    0.20
    éĥ½çŁ¥éģĵ
    0.16
    æĪijä»¬çŁ¥éģĵ
    0.16
    ä»ĸçŁ¥éģĵ
    0.15
    å°±çŁ¥éģĵ
    0.15
    ä¹ŁçŁ¥éģĵ
    0.14
     knows
    0.14
    çŁ¥éģĵ
    0.13
    æĪijçŁ¥éģĵ
    0.12
     known
    0.11
    Activations Density 0.053%

    No Known Activations