© 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. 34-GEMMASCOPE-RES-131K
    4. 91362
    Prev
    Next
    INDEX
    Explanations

    software build deploy development features

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

    The neuron fires strongly on core software-engineering content words—particularly heavyweight nouns and technical terms (e.g. “quality,” “product,” “project,” “changes,” “system”)—highlighting domain-specific keywords in CI/CD and development contexts.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ('\\
    -0.82
    éges
    -0.76
     enfermedad
    -0.75
    TokenType
    -0.74
    μμ
    -0.71
     Nutr
    -0.71
     dientes
    -0.71
    UMENTS
    -0.70
     Königreich
    -0.70
    Acces
    -0.69
    POSITIVE LOGITS
     software
    3.36
    software
    2.63
    Software
    2.44
    软件
    2.33
     Software
    2.31
     development
    2.13
    SOFTWARE
    2.06
     SOFTWARE
    2.00
     code
    1.98
     developers
    1.91
    Activations Density 0.112%

    No Known Activations