© 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. 130799
    Prev
    Next
    INDEX
    Explanations

    alignment and stability

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

    The main thing this neuron does is detect occurrences of “align” (and related forms like alignment or misalignment) in the 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
    ↵↵
    -1.73
     to
    -1.66
     solcher
    -1.49
    .
    -1.48
    oceros
    -1.47
    eapples
    -1.46
    <bos>
    -1.45
    eliness
    -1.35
     Bewußt
    -1.34
     schwerer
    -1.34
    POSITIVE LOGITS
     bús
    1.77
    哟
    1.72
     continúas
    1.72
    垍
    1.57
    
    1.45
    ⸎
    1.45
    -¡
    1.45
    帶著
    1.45
     alarma
    1.43
     parfü
    1.43
    Activations Density 0.031%

    No Known Activations