© 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. 10-GEMMASCOPE-RES-131K
    7. 57504
    Prev
    Next
    INDEX
    Explanations

    hexadecimal and numbers

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

    The neuron fires strongly on tokens that are parts of long hexadecimal or hex‐style numeric identifiers (i.e. sequences of 0–9 and a–f characters).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    坐着
    -1.87
     veya
    -1.64
    鬟
    -1.59
     profesora
    -1.59
     três
    -1.54
    所以我
    -1.53
     chercheurs
    -1.52
     cuatro
    -1.49
     bús
    -1.48
     diverses
    -1.48
    POSITIVE LOGITS
     with
    2.09
     to
    1.74
    lgari
    1.63
    BIÉN
    1.55
    чики
    1.54
     most
    1.52
     because
    1.50
    👍
    1.45
     more
    1.45
     begin
    1.44
    Activations Density 0.027%

    No Known Activations