© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129552
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on numeric literals and constant‐style identifiers (e.g. uppercase tokens and numbers) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    sequences followed by `.` or `htm`

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     ungkap
    -1.05
     también
    -0.90
     também
    -0.89
    strongly
    -0.89
    happily
    -0.87
    に向かって
    -0.87
     颜色
    -0.86
    frequently
    -0.85
    Commenting
    -0.84
     πρ
    -0.84
    POSITIVE LOGITS
    }$.
    1.03
     hilfre
    0.80
     tatsächlich
    0.79
    \_
    0.76
     premature
    0.75
    譜
    0.74
    実現
    0.73
     einzigartigen
    0.73
     pedig
    0.72
     increase
    0.71
    Activations Density 0.004%

    No Known Activations