© 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. 34-GEMMASCOPE-RES-131K
    7. 119360
    Prev
    Next
    INDEX
    Explanations

    hope this helps

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

    The neuron detects polite closing or sign-off phrases expressing goodwill assistance (e.g. “Hope this helps”).

    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
     erzä
    -0.92
    ngan
    -0.89
    =[[
    -0.88
    =()=>{
    -0.88
     talaga
    -0.84
     Resultados
    -0.82
     muna
    -0.82
    eeeee
    -0.81
    messung
    -0.81
    lde
    -0.80
    POSITIVE LOGITS
     helps
    1.70
     helpful
    1.65
    少しでも
    1.45
     helped
    1.43
     help
    1.38
     useful
    1.33
     you
    1.25
     clarifies
    1.24
     helping
    1.20
     clarify
    1.18
    Activations Density 0.064%

    No Known Activations