© 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. 34-GEMMASCOPE-RES-131K
    7. 33347
    Prev
    Next
    INDEX
    Explanations

    foreign words and titles

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

    This neuron fires on stray non‐text or control‐character tokens (e.g. the hidden “” markers) rather than normal words.

    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
     any
    -1.28
     to
    -1.15
     all
    -1.09
     many
    -1.09
     from
    -1.05
     only
    -1.04
     those
    -1.02
     makes
    -1.00
     these
    -0.98
     an
    -0.97
    POSITIVE LOGITS
     pesante
    1.04
    шир
    1.01
    èces
    0.98
     Dinas
    0.97
     }}(
    0.96
     ancor
    0.95
     attente
    0.95
    ât
    0.95
     Waray
    0.93
    entwicklung
    0.93
    Activations Density 0.002%

    No Known Activations