© 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. 34-GEMMASCOPE-RES-131K
    7. 78676
    Prev
    Next
    INDEX
    Explanations

    asking questions, listening, probing

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

    The neuron is sensitive to methodological verbs and validation terms you’d find in academic or research‐methods text (e.g. “validate,” “verify,” “manage,” “listen,” etc.).

    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
     externes
    -0.81
     Plu
    -0.79
    nical
    -0.79
    passer
    -0.79
    せっ
    -0.78
    Plu
    -0.77
     Elek
    -0.77
    PED
    -0.77
    immune
    -0.77
    plugs
    -0.76
    POSITIVE LOGITS
     probe
    1.32
     asking
    1.19
     questions
    1.14
     Probe
    1.11
     listening
    1.11
     confirm
    1.11
     encourage
    1.10
     probes
    1.08
     clarify
    1.08
     probing
    1.07
    Activations Density 0.020%

    No Known Activations