© 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. 10-GEMMASCOPE-RES-131K
    7. 102976
    Prev
    Next
    INDEX
    Explanations

    saw, sawmill, sawing

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

    The neuron strongly activates on occurrences of the substring “saw” (whether as the noun/tool, the verb, or part of a larger token).

    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
     \
    -2.91
    ୈ
    -2.52
    ’
    -2.52
    ”
    -2.42
    ]
    -2.38
    𐍉
    -2.19
    Специ
    -2.08
    的一個
    -2.08
    噗
    -2.03
     بنات
    -1.95
    POSITIVE LOGITS
    ↵↵
    2.75
     It
    2.63
    ЛЯ
    2.19
    ropractor
    2.14
    物が
    2.11
     SAW
    2.03
    %).
    2.03
     entièrement
    2.02
     cercana
    2.02
     offensichtlich
    1.97
    Activations Density 0.006%

    No Known Activations