© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 112726
    Prev
    Next
    INDEX
    Explanations

    restatements or continuations

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

    The neuron responds strongly to the discourse adverb “again,” i.e. instances where the text repeats or emphasizes a point by using “Again.”

    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
     as
    -1.75
     the
    -1.71
    錚
    -1.65
     and
    -1.57
     on
    -1.56
     Interestingly
    -1.51
     Twelve
    -1.50
     It
    -1.49
     Sixteen
    -1.49
    )
    -1.48
    POSITIVE LOGITS
    ↵
    1.80
     they
    1.64
     yako
    1.62
    躝
    1.58
     クロ
    1.53
    垆
    1.53
    每次
    1.52
     résult
    1.49
    """
    1.48
    sometimes
    1.48
    Activations Density 0.003%

    No Known Activations