© 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. 50126
    Prev
    Next
    INDEX
    Explanations

    decline or declining

    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
    ãģıãĤĬ
    -0.11
    ters
    -0.10
    disk
    -0.10
    инг
    -0.10
    ially
    -0.09
    ted
    -0.09
    desk
    -0.09
    dl
    -0.09
    ians
    -0.09
    åħĴ
    -0.09
    POSITIVE LOGITS
    phas
    0.10
    Ñĩив
    0.10
    ary
    0.10
    ãģ¾ãģĹãģ¦
    0.09
    rys
    0.09
    azon
    0.09
    ä¹ĭä½ľ
    0.09
    ä½ĵåĬĽ
    0.09
    hardt
    0.09
    ances
    0.09
    Activations Density 0.014%

    No Known Activations