© 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. 36262
    Prev
    Next
    INDEX
    Explanations

    start of sentence after punctuation

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

    The neuron is a sentence-start detector in formal (especially academic) prose, firing strongly on common opening/discourse words like “In,” “This,” “We,” or “Here.”

    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
    同样的
    -0.92
     diverses
    -0.92
    めていた
    -0.90
    Eres
    -0.90
     anschließend
    -0.88
     exemples
    -0.88
     bombe
    -0.85
     tumba
    -0.84
    Means
    -0.84
    ᝰ
    -0.84
    POSITIVE LOGITS
     recently
    1.20
     поэтому
    1.13
     therefore
    1.12
     इसलिए
    1.05
     here
    1.03
     Therefore
    1.03
     nor
    1.02
     dlatego
    1.00
     ours
    0.99
     so
    0.96
    Activations Density 0.216%

    No Known Activations