© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 66579
    Prev
    Next
    INDEX
    Explanations

    Lingard, Allow, Q, [Ro, Int

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

    The neuron strongly activates on structural‐layout tokens—things like section headers, standalone numbers, hyphens/dashes, and other nonword markers rather than ordinary running 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_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
     äldre
    -1.05
     galite
    -1.05
    pherds
    -0.95
     sociala
    -0.94
    }{{
    -0.93
    clerView
    -0.90
    standes
    -0.90
     gemens
    -0.89
    ája
    -0.88
     Ondo
    -0.88
    POSITIVE LOGITS
     and
    1.02
    ₤
    0.88
    imantan
    0.87
    plash
    0.85
     or
    0.84
    venant
    0.83
    sufficient
    0.83
    нье
    0.82
    uver
    0.81
    uesa
    0.81
    Activations Density 0.000%

    No Known Activations