© 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. 23960
    Prev
    Next
    INDEX
    Explanations

    sabotage, extortion

    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
    æ¯ĴæĢ§
    -0.09
    éħİ
    -0.09
    uce
    -0.09
    anka
    -0.08
    agma
    -0.08
     пÑĢавилÑĮно
    -0.08
    bies
    -0.08
    rise
    -0.08
    Ñij
    -0.08
    ategy
    -0.08
    POSITIVE LOGITS
    eer
    0.11
    èĢħçļĦ
    0.10
    çĪ±æĥħ
    0.10
    gable
    0.10
    èĢħ
    0.09
    alent
    0.09
    ick
    0.09
    ç©ij
    0.09
     ladder
    0.08
    è¦ģçŁ¥éģĵ
    0.08
    Activations Density 0.006%

    No Known Activations