© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 3031
    Prev
    Next
    INDEX
    Explanations

    safe [related 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
    Negative Logits
    naire
    -0.14
    UNCH
    -0.10
    /popper
    -0.10
    éĥ¨ä»½
    -0.09
     importantly
    -0.09
    ss
    -0.09
    unch
    -0.09
    ited
    -0.09
    naires
    -0.09
    cano
    -0.09
    POSITIVE LOGITS
    keeping
    0.14
    keepers
    0.12
    eners
    0.11
    -looking
    0.11
    -ish
    0.11
    idious
    0.10
    度è¿ĩ
    0.10
    æľīåºıæİ¨è¿Ľ
    0.10
    ä¸İåIJ¦
    0.10
    åı¯éĿłçļĦ
    0.10
    Activations Density 0.059%

    No Known Activations