© 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. 22-GEMMASCOPE-RES-131K
    4. 84513
    Prev
    Next
    INDEX
    Explanations

    relationship and conversation topics

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

    This neuron detects temporal and narrative‐progression markers—words and phrases that signal shifts in time or introduce the next event in a story.

    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
     of
    -1.38
    ଶ
    -1.16
    *
    -1.08
     in
    -1.05
     alongside
    -1.05
     to
    -1.04
    与
    -1.02
     systemctl
    -1.02
     are
    -1.01
    5
    -0.99
    POSITIVE LOGITS
     mutual
    1.73
    お互
    1.70
    mutual
    1.61
     mutually
    1.59
    Köszönöm
    1.44
    Mutual
    1.43
    alucía
    1.41
    Szia
    1.38
    chemia
    1.38
     between
    1.38
    Activations Density 0.026%

    No Known Activations