© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 38847
    Prev
    Next
    INDEX
    Explanations

    titles of articles and books

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

    The neuron flags standalone numeric tokens (in particular digits and multi‐digit numbers like years).

    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
     etc
    -1.18
    ne
    -1.18
    red
    -1.12
    under
    -1.09
    enn
    -1.09
    med
    -1.05
     твоя
    -1.04
    B
    -1.02
     vatten
    -1.02
    ah
    -1.01
    POSITIVE LOGITS
     dealloc
    1.34
     by
    1.25
    Sometimes
    1.15
     agarre
    1.14
     appelez
    1.13
     Rabatt
    1.11
     beschäf
    1.11
     stratégique
    1.10
     today
    1.09
     Sometimes
    1.08
    Activations Density 0.344%

    No Known Activations