© 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. 34-GEMMASCOPE-RES-131K
    7. 116149
    Prev
    Next
    INDEX
    Explanations

    If, copyright, software

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

    The neuron detects language that directly addresses or guides the reader—e.g. second-person pronouns (“you”), directive phrases (“here are,” “looking for”), and related instructional wording.

    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
    afka
    -0.90
    絣
    -0.83
     respira
    -0.76
    ربعة
    -0.76
     не
    -0.75
    unanje
    -0.72
     зем
    -0.72
    iao
    -0.71
     Petitioner
    -0.71
     suspe
    -0.70
    POSITIVE LOGITS
     Adams
    0.91
     mentality
    0.87
    Adams
    0.80
     вчера
    0.79
     Nord
    0.78
     Barclay
    0.76
     Amber
    0.75
     Flasche
    0.74
     Jugendlichen
    0.74
    section
    0.73
    Activations Density 0.003%

    No Known Activations