© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 4172
    Prev
    Next
    INDEX
    Explanations

    TWD characters and Alexandria

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

    This neuron marks the first few “content” words of a new sentence or block (e.g. the initial substantive words in a paragraph).

    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
     caffeine
    -0.78
     paws
    -0.78
    辄
    -0.78
    猥
    -0.78
    Ys
    -0.76
     snug
    -0.76
    RIO
    -0.75
    beak
    -0.75
    lask
    -0.74
    Marble
    -0.74
    POSITIVE LOGITS
     walker
    1.30
     walkers
    1.29
     Alexand
    1.11
     Alexandria
    1.06
     communities
    1.06
     Judith
    1.04
     Whisper
    0.96
     Communities
    0.96
     Daryl
    0.94
     Hers
    0.93
    Activations Density 0.007%

    No Known Activations