© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 15741
    Prev
    Next
    INDEX
    Explanations

    making it easier to compensate

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

    The neuron fires on words and phrases that denote offsetting, compensating, or counteracting something.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_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
    ucid
    0.62
    ulton
    0.56
     সাই
    0.56
     repeatable
    0.54
    orul
    0.54
    ዴት
    0.54
    拜
    0.54
    subdir
    0.54
    𝘤
    0.54
     தின
    0.53
    POSITIVE LOGITS
     compensate
    2.20
     compensates
    2.13
     компен
    2.05
     compensating
    1.91
     compens
    1.88
    compens
    1.85
     compensated
    1.79
     Compens
    1.77
    补偿
    1.71
     compensation
    1.70
    Activations Density 0.403%

    No Known Activations