© 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. 22-GEMMASCOPE-RES-131K
    7. 130298
    Prev
    Next
    INDEX
    Explanations

    The main thing this neuron does is detect verbs that describe quantitative changes or improvements (e.g., increase, improve, enhance, reduce).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    improve or increase success

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    Not in Any Lists

    No Comments

    Negative Logits
     зді
    -1.53
    廊下
    -1.38
     Thankfully
    -1.31
     yeterli
    -1.26
    しては
    -1.23
    始めて
    -1.23
    人には
    -1.22
     спеціаль
    -1.21
    了一会儿
    -1.21
     Plenty
    -1.20
    POSITIVE LOGITS
     the
    3.05
     their
    1.59
     chances
    1.50
     its
    1.50
     ktoré
    1.28
     компания
    1.26
     partecipanti
    1.23
     our
    1.19
     opportunities
    1.17
    bürgermeister
    1.17
    Activations Density 0.084%

    No Known Activations