© 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. 41085
    Prev
    Next
    INDEX
    Explanations

    speaking to

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

    The neuron detects instances of the verb “speak” (in various tenses/forms) immediately followed by “to,” i.e. constructions like “speaking to,” “spoke to,” or “speaks to.”

    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
     Both
    -2.30
     Countries
    -2.02
    一会
    -1.97
    線が
    -1.94
    ᯾
    -1.91
    LEBR
    -1.87
    蔼
    -1.85
    .
    -1.84
     swivel
    -1.79
     zowel
    -1.78
    POSITIVE LOGITS
    ⓷
    2.25
    皤
    2.08
    hasilkan
    1.95
    谖
    1.90
    欵
    1.86
     giapponese
    1.85
     femminile
    1.84
     ſch
    1.81
    充满了
    1.80
     penghar
    1.80
    Activations Density 0.013%

    No Known Activations