© 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. 36410
    Prev
    Next
    INDEX
    Explanations

    continues or terminates

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

    The neuron activates on words describing pronounced changes in magnitude or position—terms like “drop,” “off,” “rises,” or “steeply.”

    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
     soggior
    -1.11
    Neden
    -1.10
     pesquis
    -1.09
    もあった
    -1.08
    Çok
    -1.06
     cucchia
    -1.05
     asciug
    -1.02
     muligt
    -1.01
     jednoduch
    -1.01
    zyż
    -1.01
    POSITIVE LOGITS
     at
    1.34
     into
    1.30
     from
    1.28
     beyond
    1.14
     continuation
    1.01
     extends
    0.98
    G
    0.97
     vanaf
    0.94
     hingga
    0.94
     continues
    0.94
    Activations Density 0.060%

    No Known Activations