© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 13517
    Prev
    Next
    INDEX
    Explanations

    decreased, low, or missing values

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

    The neuron activates on comparative terms or adjectives/adverbs indicating a quantitative comparison (e.g. “smaller,” “greater,” “lower,” “increased,” “poorer,” etc.).

    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
    tonsoft
    -1.60
    haviours
    -1.55
    ícias
    -1.53
    mdash
    -1.50
     senza
    -1.48
    raeli
    -1.46
    了一会儿
    -1.45
    izando
    -1.41
    ctomy
    -1.39
    kenen
    -1.39
    POSITIVE LOGITS
     access
    1.70
     use
    1.49
     overall
    1.44
     operational
    1.35
     extensive
    1.35
     established
    1.34
     ability
    1.30
     thorough
    1.30
     used
    1.29
    ありません
    1.29
    Activations Density 0.057%

    No Known Activations