© 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. 104055
    Prev
    Next
    INDEX
    Explanations

    if clauses

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

    the word "If" at the beginning of conditional clauses or sentences.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron strongly activates on the subordinating conjunction “If” (i.e. the start of conditional clauses).

    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
     are
    -2.02
     but
    -1.89
     may
    -1.85
     It
    -1.84
    }
    -1.84
     might
    -1.80
    k
    -1.80
    x
    -1.80
     has
    -1.77
     can
    -1.75
    POSITIVE LOGITS
    rophes
    2.13
    戽
    1.90
     penyanyi
    1.88
    蟶
    1.86
    ผม
    1.69
     például
    1.68
    źdz
    1.65
    樯
    1.63
    ꒦
    1.62
     prêts
    1.61
    Activations Density 0.045%

    No Known Activations