© 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. 22-GEMMASCOPE-RES-131K
    4. 39696
    Prev
    Next
    INDEX
    Explanations

    html content type

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

    The neuron strongly activates on non‐alphanumeric code punctuation—especially closing braces, parentheses, comparison operators, and commas.

    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
     Cette
    -1.09
     Diese
    -1.09
     These
    -1.04
     ideen
    -1.02
    )++;
    -1.02
     indicando
    -1.02
    nang
    -1.01
     电视
    -0.98
    Appellants
    -0.98
    iin
    -0.96
    POSITIVE LOGITS
     when
    1.44
     in
    1.34
     there
    1.25
     because
    1.24
     after
    1.16
     debería
    1.14
     you
    1.14
     despite
    1.13
     although
    1.13
     most
    1.11
    Activations Density 0.001%

    No Known Activations