© 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. 70597
    Prev
    Next
    INDEX
    Explanations

    various scripts and languages

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

    This neuron fires on the formal, legal‐boilerplate language and numeric tokens you see in corporate filings and announcements—especially phrases around “non-binding” and “definitive agreement” wording (and the associated section numbers and dates).

    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
     like
    -1.45
     slowly
    -1.30
     τὸ
    -1.28
     chré
    -1.28
     ὁ
    -1.23
     τὴν
    -1.23
     many
    -1.20
     quickly
    -1.16
     liked
    -1.14
     włosów
    -1.13
    POSITIVE LOGITS
    ּוֹ
    2.03
    וּ
    1.79
    According
    1.47
    ῃ
    1.47
    Beskrivelse
    1.45
    ༘
    1.43
     Jerusalén
    1.42
     Suiza
    1.42
     кроссовки
    1.41
     menjalani
    1.39
    Activations Density 0.001%

    No Known Activations