© 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. 64894
    Prev
    Next
    INDEX
    Explanations

    preposition followed by specific phrase

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

    The neuron fires on formal clause-opening prepositional phrases (e.g. “In the …,” “As with …,” “Without prejudice,” “When citing,” etc.) typical of legal or policy text.

    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
     lately
    -1.14
     even
    -1.12
    gilet
    -1.06
     hierro
    -1.05
     kapas
    -1.04
     didn
    -1.04
     блюдо
    -1.04
    arons
    -1.02
     now
    -1.01
    がある
    -1.00
    POSITIVE LOGITS
     Fundação
    1.16
    ṇa
    1.16
    isling
    1.16
    Occasionally
    1.15
    gether
    1.15
    ifrance
    1.14
    водом
    1.14
     Haye
    1.12
    Each
    1.10
    1.09
    Activations Density 0.046%
    No Known Activations