© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-14B
    5. 20-RESID-BATCHTOPK-65K__L0-80
    7. 15159
    Prev
    Next
    INDEX
    Explanations

    Fifth Avenue Shoe Repair

    np_acts-logits-general · gemini-2.5-flash-lite

    quotations or quoted speech, particularly detecting the beginning or end of direct quotes and dialogue.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jrosseruk
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    adamkarvonen/qwen3-14b-saes/saes_Qwen_Qwen3-14B_batch_top_k/resid_post_layer_20
    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
    kest
    -0.24
    INI
    -0.20
    æľīèĩªå·±çļĦ
    -0.20
    _kind
    -0.19
    cke
    -0.19
    åīĸ
    -0.18
    åĢĶ
    -0.18
    allee
    -0.18
    ested
    -0.17
    éĿ³
    -0.17
    POSITIVE LOGITS
    åıŁ
    0.23
     fictional
    0.20
     Constit
    0.19
    ç¼Ł
    0.19
    æĬĢæľ¯æ°´å¹³
    0.19
    ç®ĹäºĨ
    0.19
    zo
    0.19
    åºŁè¯Ŀ
    0.18
     hypo
    0.18
     MICRO
    0.18
    Activations Density 0.526%

    No Known Activations