© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 32773
    Prev
    Next
    INDEX
    Explanations

    instead or rather

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

    This neuron detects first-person or conversational discourse markers expressing intent, opinion, or deference (e.g. “I’ll,” “I think,” “we’ll leave it to X,” “let’s,” “should,” etc.).

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    穿搭
    -0.88
     Field
    -0.86
    棂
    -0.86
     focus
    -0.85
    cVar
    -0.84
    ⍟
    -0.84
    |()
    -0.81
     қара
    -0.80
     undergrad
    -0.80
    acão
    -0.79
    POSITIVE LOGITS
     rather
    1.13
    rather
    0.96
     instead
    0.95
     description
    0.95
     explanation
    0.90
     voordelen
    0.88
     بهتر
    0.88
    而不是
    0.87
     descri
    0.86
     plutôt
    0.86
    Activations Density 0.032%

    No Known Activations