© 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. 22-GEMMASCOPE-RES-131K
    7. 68804
    Prev
    Next
    INDEX
    Explanations

    consequences and explanations

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

    This neuron strongly activates on causal/discourse-linking adverbs (especially “so” and similar connectors like “thus”) that introduce or link clauses.

    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
    などを
    -1.08
    <bos>
    -1.02
    ↵↵
    -0.95
    ettes
    -0.94
    かって
    -0.94
    Согласно
    -0.94
    ﷻ
    -0.93
    がって
    -0.93
    諍
    -0.91
     tabu
    -0.90
    POSITIVE LOGITS
    わけで
    1.14
    こともあり
    1.09
    為に
    1.09
    ところは
    1.08
    うまい
    1.07
    わけです
    1.05
    こいつ
    1.05
    あぁ
    1.05
    久し
    1.04
    也算
    1.04
    Activations Density 0.060%

    No Known Activations