© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 101259
    Prev
    Next
    INDEX
    Explanations

    counterfactual past scenarios

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

    This neuron strongly activates on occurrences of the phrase “would have” (i.e. the hypothetical/modal construction “would have”).

    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
    ︕
    -1.93
    茈
    -1.89
    брь
    -1.59
     peuvent
    -1.59
    
    -1.55
    镲
    -1.54
    ar
    -1.53
     fiables
    -1.52
     знают
    -1.51
     Når
    -1.49
    POSITIVE LOGITS
     Would
    1.77
     It
    1.63
    ntax
    1.61
     Someone
    1.59
     last
    1.59
    If
    1.55
     result
    1.55
    However
    1.54
     justifyContent
    1.53
    》、《
    1.48
    Activations Density 0.015%

    No Known Activations