© 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. 22-GEMMASCOPE-RES-131K
    7. 16207
    Prev
    Next
    INDEX
    Explanations

    late periods or stages

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

    The neuron is detecting numeric tokens (numbers, such as years, dates, counts, and other multi‐digit figures).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     first
    -1.20
     a
    -1.11
     seventy
    -1.07
     initial
    -1.06
    ftet
    -1.05
     første
    -1.02
     initially
    -1.01
    لهم
    -0.99
    ouvrir
    -0.98
     initializes
    -0.98
    POSITIVE LOGITS
    late
    1.27
     quando
    1.09
     late
    1.09
     יותר
    1.07
     akhir
    1.05
     Late
    1.02
    digans
    0.99
    рих
    0.97
     azért
    0.96
     inkább
    0.96
    Activations Density 0.044%

    No Known Activations