© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-32B
    5. 32-RESID-BATCHTOPK-65K
    7. 52320
    Prev
    Next
    INDEX
    Explanations

    "tr"/"tra"/"tram"/"intr" prefixes followed by medical/scientific terms

    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
    -0.11
    berra
    -0.10
    ëģĶ
    -0.10
    uras
    -0.10
    oir
    -0.10
    à²ł
    -0.09
    eds
    -0.09
    RIPT
    -0.09
    GING
    -0.09
    ibern
    -0.09
    POSITIVE LOGITS
    duct
    0.11
    most
    0.11
    ernal
    0.10
    ãģĹãģ¾ãģĨ
    0.10
    insics
    0.10
    lude
    0.09
    ly
    0.09
    acies
    0.09
    ãģħ
    0.09
    eger
    0.09
    Activations Density 0.074%

    No Known Activations