© 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. 23598
    Prev
    Next
    INDEX
    Explanations

    warned of dangers or risks

    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
    妾
    -0.10
    æĴĴ
    -0.10
    æĺ¯æľĢ好çļĦ
    -0.10
    ORMAL
    -0.09
    ew
    -0.09
    éļĶ
    -0.09
     Corpus
    -0.09
    座è°Ī
    -0.08
    éª
    -0.08
    å¨ĥ
    -0.08
    POSITIVE LOGITS
    èѦåijĬ
    0.14
     warns
    0.12
     dangers
    0.11
     warned
    0.11
     warnings
    0.10
     cautioned
    0.10
     warning
    0.10
     Warning
    0.10
     risks
    0.10
    âļł
    0.09
    Activations Density 0.029%

    No Known Activations