© 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. 22-GEMMASCOPE-RES-131K
    7. 73953
    Prev
    Next
    INDEX
    Explanations

    speaking, thinking, predicting, intending

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

    This neuron fires on reporting or commentary verbs (like described, think, predicted, suggests) and their immediate “that/it” constructions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    躯
    -0.98
    ąg
    -0.85
    本の
    -0.85
     concepción
    -0.84
    ონ
    -0.84
     издание
    -0.81
     motivación
    -0.78
    kset
    -0.78
    Schneider
    -0.78
     conoz
    -0.77
    POSITIVE LOGITS
    rů
    1.01
     anteriormente
    0.95
     promised
    0.94
     earlier
    0.93
    orgio
    0.92
    대로
    0.91
     them
    0.88
    bove
    0.87
     warned
    0.87
     personnaliser
    0.87
    Activations Density 0.044%

    No Known Activations