© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 85658
    Prev
    Next
    INDEX
    Explanations

    retrograde direction or atrial events

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

    The neuron identifies technical terms related to directional flow, particularly "retrograde" and variations of "antegrade" along with Chinese names containing "shun" or "sun".

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

    The neuron responds to directional “-grade” terms, specifically spotting occurrences of words like “retrograde” or “anterograde.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    0
    -3.31
    ('');
    -2.59
     mutta
    -2.34
     факторов
    -2.28
     sportowe
    -2.23
    簠
    -2.17
    並不
    -2.17
    頃に
    -2.16
     pero
    -2.13
     of
    -2.11
    POSITIVE LOGITS
    '
    2.86
    me
    2.58
    鲼
    2.52
    ጺ
    2.48
    媜
    2.41
    婓
    2.38
    ll
    2.28
    Ꮄ
    2.27
    你
    2.27
    aren
    2.27
    Activations Density 0.003%

    No Known Activations