© 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. 116489
    Prev
    Next
    INDEX
    Explanations

    telling the truth

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

    The neuron detects explicit references to truthfulness and deception—words like “telling,” “truth,” “lie,” and “realize.”

    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
    preuve
    -0.87
    |^{
    -0.81
     ciudadano
    -0.80
     persuasive
    -0.79
    lık
    -0.79
    Images
    -0.75
     models
    -0.75
     a
    -0.75
     nossas
    -0.75
     display
    -0.74
    POSITIVE LOGITS
     truth
    4.31
     Truth
    3.41
    truth
    3.27
    Truth
    3.22
     TRUTH
    2.89
     truths
    2.58
     verdad
    2.23
     verità
    2.17
    真相
    2.16
     truthful
    2.06
    Activations Density 0.031%

    No Known Activations