© 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. 35164
    Prev
    Next
    INDEX
    Explanations

    testing suites

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

    The neuron strongly activates on numeric literals (tokens containing digits or numbers) within code.

    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
     when
    -1.19
     if
    -0.98
     to
    -0.92
    しますが
    -0.85
    していますが
    -0.85
     only
    -0.85
     zü
    -0.85
    しているので
    -0.84
     from
    -0.83
     with
    -0.82
    POSITIVE LOGITS
    RegressionTest
    1.15
     abaixo
    1.07
    對於
    1.05
     mantenga
    1.03
    خبر
    1.02
     Relic
    1.02
    今まで
    1.00
    TestBed
    1.00
    下記
    0.97
    对于
    0.96
    Activations Density 0.007%

    No Known Activations