© 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. 73691
    Prev
    Next
    INDEX
    Explanations

    code and markup delimiters

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

    The neuron is detecting structural formatting tokens—especially long runs of whitespace (indentation), newlines, braces, and quotes that mark code or XML layout.

    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
    erweise
    -1.24
     treball
    -1.15
    orul
    -1.14
     كمان
    -1.09
     LEYENDO
    -1.09
    享受
    -1.09
     लेकिन
    -1.03
     prescriptions
    -1.02
     plutôt
    -1.02
     types
    -1.02
    POSITIVE LOGITS
    assemblée
    1.10
     gantung
    1.08
    Bises
    1.02
    𝒙
    1.02
    詫
    0.99
     its
    0.98
     Pillars
    0.98
     giacca
    0.98
    whereas
    0.98
    Crédit
    0.96
    Activations Density 0.002%

    No Known Activations