© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 26378
    Prev
    Next
    INDEX
    Explanations

    great success

    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
    que
    -0.09
    ette
    -0.09
    yla
    -0.08
    ãģ¹ãģį
    -0.08
    äºĨ大éĩıçļĦ
    -0.08
    ester
    -0.08
    aget
    -0.08
    9
    -0.08
     Gamb
    -0.08
    trap
    -0.08
    POSITIVE LOGITS
    -grand
    0.18
    fully
    0.13
    ëģĶ
    0.12
    sword
    0.11
    rix
    0.11
    éĺµå®¹
    0.10
    ÃŃsimo
    0.10
    coat
    0.10
    ä¸İåIJ¦
    0.10
     havoc
    0.10
    Activations Density 0.049%

    No Known Activations