© 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. 10-GEMMASCOPE-RES-131K
    4. 9614
    Prev
    Next
    INDEX
    Explanations

    verbs followed by other words

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

    The neuron fires on verbs that announce or describe key actions or methodological steps (often with first‐person or organizational subjects), such as “were taken,” “created,” “will offer,” “we test,” “she started,” 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_10/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
     therefore
    -1.43
     due
    -1.41
     όμως
    -1.41
     problemática
    -1.41
    由於
    -1.35
    preuve
    -1.30
    五六
    -1.30
     thus
    -1.20
     Это
    -1.16
    由于
    -1.16
    POSITIVE LOGITS
    襌
    1.38
     ajedrez
    1.38
     hè
    1.27
    越来越
    1.27
     today
    1.26
    ”
    1.23
     julio
    1.22
    ations
    1.19
    Charsets
    1.18
     diseñadores
    1.18
    Activations Density 0.198%

    No Known Activations