© 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. 38829
    Prev
    Next
    INDEX
    Explanations

    stuck or falling words

    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
    ç®ĢåİĨ
    -0.10
    èĩªè´Ł
    -0.09
    梦å¢ĥ
    -0.09
    诤
    -0.09
    uce
    -0.08
    -facing
    -0.08
    uzzi
    -0.08
    å°ı鸣
    -0.08
    nze
    -0.08
    -rich
    -0.08
    POSITIVE LOGITS
    STA
    0.10
    gary
    0.10
    èĥĽ
    0.10
    ailed
    0.09
    weg
    0.09
    rig
    0.09
     Ryu
    0.08
    èĮĨ
    0.08
    ry
    0.08
    ä¸ĢéĹª
    0.08
    Activations Density 0.007%

    No Known Activations