© 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. 13195
    Prev
    Next
    INDEX
    Explanations

    There is/are noun phrases

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

    The neuron is triggered by existential or indefinite noun phrases—basically “there is/are…” or standalone “a/an” constructions signaling “there exists” something.

    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
     having
    -1.44
     its
    -1.30
     svoj
    -1.23
     ayant
    -1.16
     swoich
    -1.16
    ratislava
    -1.13
     svých
    -1.11
    openha
    -1.09
    我们
    -1.06
    celotti
    -1.04
    POSITIVE LOGITS
     there
    1.14
    <bos>
    1.06
    関連リンク
    1.05
     automne
    1.02
     no
    1.02
    owels
    1.00
     misconception
    0.96
    धान
    0.96
     actualité
    0.96
     bambou
    0.95
    Activations Density 0.383%

    No Known Activations