© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 84915
    Prev
    Next
    INDEX
    Explanations

    by the way

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

    The neuron fires on shorthand internet acronyms—especially “btw” (“by the way”)—i.e. casual initialisms used in informal text.

    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
    </td>
    -3.39
    呖
    -2.83
    䉴
    -2.53
    }
    -2.50
    鉝
    -2.44
    䊐
    -2.39
    </i>
    -2.33
    昻
    -2.30
    斟
    -2.28
    ↵↵
    -2.25
    POSITIVE LOGITS
    t
    2.61
    垆
    2.30
     However
    2.30
    鬭
    2.28
     immobili
    2.22
    СКИ
    2.20
    榈
    2.17
    身后的
    2.11
    y
    2.11
     generó
    2.08
    Activations Density 0.001%

    No Known Activations