© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72255
    Prev
    Next
    INDEX
    Explanations

    scientific and legal concepts

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

    The neuron fires strongly on domain-specific, multi-syllable noun keywords (often technical or scientific terms) that typically head or label sections—e.g. “histograms,” “tanks,” “cytotoxicity,” “therapies,” “signals,” etc.

    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
     or
    -1.28
     If
    -1.09
     for
    -1.09
     their
    -1.08
     L
    -1.03
    NF
    -1.02
     with
    -1.02
     if
    -1.02
     k
    -0.98
    ........
    -0.98
    POSITIVE LOGITS
     laſt
    1.28
    samt
    1.20
     medlemmer
    1.18
    鸲
    1.15
     tika
    1.15
    Ressources
    1.14
     persil
    1.11
     ホルダー
    1.10
     sekä
    1.09
    neth
    1.09
    Activations Density 0.095%

    No Known Activations