© 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. 34-GEMMASCOPE-RES-131K
    7. 32407
    Prev
    Next
    INDEX
    Explanations

    improvement and getting better

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

    The neuron is primarily triggering on occurrences of the verb “to be” (e.g. “is,” “to be,” “be”) in various contexts.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    Aktual
    -1.09
    ","","
    -1.04
     mereka
    -1.02
     atrak
    -1.01
     penser
    -0.99
     agir
    -0.99
    」,
    -0.98
     récon
    -0.98
    ungkap
    -0.98
    给他们
    -0.96
    POSITIVE LOGITS
     improvements
    1.83
     improved
    1.76
     improve
    1.76
     improvement
    1.74
     improving
    1.46
     offseason
    1.41
     maturity
    1.38
     Improved
    1.33
     consistency
    1.28
     become
    1.27
    Activations Density 0.010%

    No Known Activations