© 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. 128948
    Prev
    Next
    INDEX
    Explanations

    The neuron consistently lights up on future-tense constructions—especially the “going to be/have …” pattern signaling a prediction or upcoming event.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    going to be

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     obecnie
    -1.38
     will
    -1.37
     telah
    -1.35
    这是什么
    -1.30
    还会
    -1.30
     unangemessene
    -1.28
    R
    -1.27
     recently
    -1.26
     would
    -1.23
     can
    -1.21
    POSITIVE LOGITS
     fiabilité
    1.18
     vinaigre
    1.16
     a
    1.16
     rechange
    1.16
     prestation
    1.13
     matériau
    1.13
    📘
    1.12
     prochain
    1.11
     maquillage
    1.10
    ㈠
    1.09
    Activations Density 0.018%

    No Known Activations