© 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. 10-GEMMASCOPE-RES-131K
    4. 104750
    Prev
    Next
    INDEX
    Explanations

    above mentioned items

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

    This neuron detects reference phrases that point back to earlier text, especially “above” plus a noun (e.g. “the above statement/problem/situation/cases/link”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     both
    -1.87
     when
    -1.79
     during
    -1.66
     especially
    -1.55
     like
    -1.54
     as
    -1.52
     throughout
    -1.43
     across
    -1.41
     this
    -1.35
     after
    -1.35
    POSITIVE LOGITS
    Above
    1.61
    above
    1.55
     acima
    1.42
     foregoing
    1.41
    以上
    1.40
    mentioned
    1.36
    5
    1.35
     aforementioned
    1.33
    The
    1.29
    celotti
    1.27
    Activations Density 0.046%

    No Known Activations