© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25799
    Prev
    Next
    INDEX
    Explanations

    finding boundaries and constraints

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

    This neuron detects methodological or procedural future‐oriented verbs (e.g. “will,” “plan to,” “carry out,” “further,” “localize,” “elucidate”) describing planned actions in research contexts.

    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
     heilige
    -1.01
    😯
    -0.97
    citi
    -0.96
    Adaptor
    -0.96
     recens
    -0.94
    spiele
    -0.92
     pittores
    -0.90
    sonder
    -0.90
    💌
    -0.90
     subir
    -0.89
    POSITIVE LOGITS
     confirmed
    1.30
     information
    1.26
     certainty
    1.16
     known
    1.13
     evidence
    1.11
     bestätigen
    1.10
     restricciones
    1.02
     narrowing
    1.01
     based
    0.99
     eliminated
    0.98
    Activations Density 0.054%

    No Known Activations