© 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. 82696
    Prev
    Next
    INDEX
    Explanations

    video and temporal concepts

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

    The neuron activates on numeric literals in code, especially the multi‐digit parameter values used in layer definitions (e.g. kernel sizes, filter counts, strides).

    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
     development
    -1.00
    偕
    -0.78
    development
    -0.77
     sworn
    -0.77
     Development
    -0.75
    사진
    -0.74
    ราบ
    -0.69
     dev
    -0.68
    omock
    -0.68
     Saif
    -0.68
    POSITIVE LOGITS
     video
    1.45
    Video
    1.24
     videos
    1.23
     Video
    1.23
    video
    1.23
    Temporal
    1.23
    temporal
    1.21
     Temporal
    1.19
    视频
    1.15
     temporal
    1.15
    Activations Density 0.013%

    No Known Activations