© 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
    2. Gemma-3-27B
    3. 31-GEMMASCOPE-2-RES-262K
    4. 189280
    Prev
    Next
    INDEX
    Explanations

    two peopleThis is a simple, concise explanation within the word count (2 words).It finds a strong pattern: "two" from `MAX_ACTIVATING_TOKENS` is often followed by nouns representing countable entities like "people", "levels", "pairs", suggesting the neuron might attend to quantities or dual states/entities. "two people" specifically appears in the activating texts and fits the context. two people

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

    exploring and examining

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-pt/resid_post/layer_31_width_262k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     on
    0.51
     with
    0.50
     for
    0.50
     doubts
    0.49
     additional
    0.47
     no
    0.47
     contingencies
    0.47
     glimpse
    0.45
     it
    0.45
     mains
    0.45
    POSITIVE LOGITS
    atthakath
    0.59
    athor
    0.55
    aptop
    0.53
    arantad
    0.53
    attice
    0.52
     मेथ
    0.52
    illings
    0.51
    龋
    0.50
     हेक्टेयर
    0.50
    uraea
    0.49
    Activations Density 0.000%

    No Known Activations