© 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. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 89754
    Prev
    Next
    INDEX
    Explanations

    medical suffixes and roots

    np_acts-logits-general · gemini-2.5-flash-lite

    medical and anatomical terms, particularly those related to body parts and medical conditions.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron is looking for common mid-word character n-grams (four- to six-letter substrings) regardless of overall word boundaries.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     ważne
    -1.03
     several
    -0.97
    อย่าง
    -0.96
     perhaps
    -0.91
     those
    -0.81
     countless
    -0.79
     another
    -0.78
    ább
    -0.77
     arguably
    -0.77
     primary
    -0.75
    POSITIVE LOGITS
    ぇ
    1.25
    except
    0.90
     الصد
    0.86
     perman
    0.82
    milliseconds
    0.81
    ందు
    0.79
    yscale
    0.79
    ypress
    0.79
    ſs
    0.77
    eigen
    0.77
    Activations Density 0.096%

    No Known Activations