© 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. 54889
    Prev
    Next
    INDEX
    Explanations

    title: followed by text

    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
    hat
    -0.08
     pitch
    -0.08
    ?v
    -0.08
    ré
    -0.08
    adies
    -0.08
    rig
    -0.08
    eno
    -0.08
    es
    -0.08
    enez
    -0.08
    bat
    -0.07
    POSITIVE LOGITS
     ÑĥÑĩеб
    0.11
    å»ĵ
    0.11
    aeper
    0.11
     wida
    0.10
    å°ijè§ģ
    0.10
    _ACK
    0.09
    ÑģпоÑģоб
    0.09
    ì´ī
    0.09
    aina
    0.09
    æľ¨åľ°æĿ¿
    0.09
    Activations Density 0.031%

    No Known Activations