© 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. Qwen3-14B
    5. 20-RESID-BATCHTOPK-65K__L0-80
    7. 13750
    Prev
    Next
    INDEX
    Explanations

    deception, hiding, or cheating

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    硼
    -0.21
    overn
    -0.20
    bak
    -0.20
    æĸ¥
    -0.20
    wright
    -0.20
    indi
    -0.20
    天èµĭ
    -0.19
    ãĤĴéĽĨ
    -0.19
     nuis
    -0.19
    vig
    -0.19
    POSITIVE LOGITS
    éĢĥéģ¿
    0.31
    éļIJçŀĴ
    0.30
    躲
    0.28
    éĢĥ
    0.28
    éĢĥèĦ±
    0.27
    -hide
    0.27
     hiding
    0.27
    åģ·åģ·
    0.27
    æĵį纵
    0.26
    éĢłåģĩ
    0.26
    Activations Density 0.146%

    No Known Activations