© 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. 1641
    Prev
    Next
    INDEX
    Explanations

    problem, gap, wage, award, conclusion

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

    The neuron activates on the single most central “topic” word in a sentence—typically the main noun or verb that carries the core meaning.

    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
    </b>
    -1.59
    "]:
    -1.46
    rounder
    -1.41
    </h2>
    -1.39
     }=\
    -1.34
    eacute
    -1.32
    SetColor
    -1.23
    TokenNameDOT
    -1.23
    </h1>
    -1.21
    illoin
    -1.20
    POSITIVE LOGITS
     of
    1.52
    ことがあります
    1.47
    となっています
    1.46
     our
    1.45
     dimaksud
    1.44
     Wednesday
    1.42
    ⃕
    1.36
     its
    1.34
     mysteriously
    1.33
     has
    1.31
    Activations Density 18.589%

    No Known Activations