© 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. 10-GEMMASCOPE-RES-131K
    7. 124592
    Prev
    Next
    INDEX
    Explanations

    FrmCounter, FrmFont, Mike Elliott, Gold Dust

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

    The neuron strongly activates on the substring “frm” (in tokens like “Frm” or “frm”), i.e. identifiers or class names containing that sequence.

    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
    }
    -3.19
    ↵
    -3.00
     \&
    -2.94
     که
    -2.61
    \\
    -2.48
     nordrhein
    -2.44
     что
    -2.34
     empathetic
    -2.34
     objections
    -2.33
    戩
    -2.33
    POSITIVE LOGITS
    埽
    2.97
    虓
    2.95
    幛
    2.89
    塒
    2.73
    _____________
    2.66
    
    2.55
     dinam
    2.53
    癭
    2.41
    哢
    2.39
    鰨
    2.36
    Activations Density 0.000%

    No Known Activations