© 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. 22-GEMMASCOPE-RES-131K
    7. 70425
    Prev
    Next
    INDEX
    Explanations

    good intentions sincere effort

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

    This neuron detects phrases evaluating or highlighting someone’s intentions—especially positive intent expressions like “well-meaning” or “best intentions.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    prende
    -0.94
    ̏
    -0.94
     Resili
    -0.93
     now
    -0.93
     of
    -0.91
    Resili
    -0.90
    RotationOrder
    -0.88
    ได้
    -0.87
     DialogInterface
    -0.87
    izace
    -0.86
    POSITIVE LOGITS
     intentions
    1.45
     trying
    1.31
     proberen
    1.24
     sincere
    1.20
    intent
    1.16
     пыта
    1.12
     však
    1.05
     prób
    1.05
     intention
    1.04
     TRY
    1.03
    Activations Density 0.020%

    No Known Activations