© 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. 25984
    Prev
    Next
    INDEX
    Explanations

    sequence continuation words

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

    The neuron is primarily responding to structural or formatting cues—punctuation marks, line breaks, and other non‐word tokens.

    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
     sauvage
    -0.90
     virtuel
    -0.88
     maa
    -0.87
    wind
    -0.84
    vasive
    -0.82
     minimalista
    -0.82
     fondamental
    -0.82
    lty
    -0.81
     hè
    -0.81
     dévelo
    -0.78
    POSITIVE LOGITS
    今回も
    1.55
     now
    1.41
     precedent
    1.32
     history
    1.23
     will
    1.21
     сейчас
    1.20
     again
    1.17
     hope
    1.13
     so
    1.13
     diesmal
    1.12
    Activations Density 0.030%

    No Known Activations