© 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. 18843
    Prev
    Next
    INDEX
    Explanations

    level

    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
    _eof
    -0.10
    以åıĬåħ¶ä»ĸ
    -0.09
    è±Ĩ
    -0.09
    ä¸ŃåĽ½æĸĩåĮĸ
    -0.08
    ä¹±
    -0.08
    è¶´åľ¨
    -0.08
    ãģ»ãģĨ
    -0.08
     dáng
    -0.08
    两大
    -0.08
    åıijçİ°èĩªå·±
    -0.08
    POSITIVE LOGITS
     level
    0.25
    level
    0.23
    (level
    0.21
    ãĥ¬ãĥĻãĥ«
    0.21
    å±Ĥ次
    0.21
     Level
    0.21
     levels
    0.21
    Level
    0.20
    -level
    0.20
     nivel
    0.19
    Activations Density 0.159%

    No Known Activations