© 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. 17046
    Prev
    Next
    INDEX
    Explanations

    words ending in able, ish, ulent

    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
    ding
    -0.09
    \Builder
    -0.09
    quets
    -0.09
    ðŁı»
    -0.09
    å²ĺ
    -0.08
    æľ¬ç«Ļ
    -0.08
    åIJ±
    -0.08
     surv
    -0.08
    ders
    -0.08
     bore
    -0.08
    POSITIVE LOGITS
    able
    0.09
    Ĭ
    0.09
    tron
    0.09
    omm
    0.08
     اÙĦأس
    0.08
    gram
    0.08
    ãĥ¬ãĥ¼ãĤ·ãĥ§ãĥ³
    0.08
    è·ĸ
    0.08
    ability
    0.08
    æ±ł
    0.08
    Activations Density 0.002%

    No Known Activations