© 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. 3695
    Prev
    Next
    INDEX
    Explanations

    collected data and performed assessments

    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
    INO
    -0.09
    çĶļèĩ³æĺ¯
    -0.09
    arine
    -0.08
    nic
    -0.08
    åĺĺ
    -0.08
    ä¸Ģ个好çļĦ
    -0.08
     mort
    -0.08
    éĻ©
    -0.08
    ALLOC
    -0.08
    ì¡°
    -0.08
    POSITIVE LOGITS
     ÙĪÙĥاÙĨت
    0.09
     identical
    0.09
    åĪĨ为
    0.08
    ãģĮãģĤãĤĬãģ¾ãģĹãģŁ
    0.08
    0.08
    å¦Ĥä¸ĭ
    0.08
     два
    0.08
    á»Ļc
    0.07
    çĦ±
    0.07
     chose
    0.07
    Activations Density 2.916%

    No Known Activations