© 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. 84908
    Prev
    Next
    INDEX
    Explanations

    it is useful

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

    The neuron detects common academic or expository “hedging” or commentary constructions (e.g. “it might be…”, “it is useful to…”) that introduce a claim or observation.

    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
    aware
    -0.90
    Wise
    -0.88
    izada
    -0.84
    finanz
    -0.84
     unwise
    -0.82
    一款
    -0.82
    娴
    -0.81
    ₺
    -0.80
    wise
    -0.80
     cabe
    -0.79
    POSITIVE LOGITS
     instructive
    1.23
     convenient
    1.23
     useful
    1.19
     vhodné
    0.96
     удоб
    0.94
    不妨
    0.93
     Convenient
    0.93
     útil
    0.93
     interesting
    0.90
     Useful
    0.90
    Activations Density 0.060%

    No Known Activations