© 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. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 2214
    Prev
    Next
    INDEX
    Explanations

    scientific terms and concepts related to chemistry, biology, and quantum physics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    scientific notation, chemical symbols, and statistical data points.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    This neuron detects scientific or technical terms in parentheses, particularly chemical formulas, abbreviations and units.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     للاسماء
    -0.70
     الحره
    -0.60
    AutoModerator
    -0.56
    الحياه
    -0.55
     OFDb
    -0.55
    RotationOrder
    -0.54
    fromnode
    -0.54
    OGND
    -0.52
    Externé
    -0.50
    ConverterFactory
    -0.50
    POSITIVE LOGITS
     pinulongan
    0.39
     kasarigan
    0.37
    INTERFACE
    0.37
     mesmo
    0.36
     Same
    0.36
     same
    0.36
     sele
    0.35
     således
    0.34
     Guten
    0.34
     proceder
    0.34
    Activations Density 0.405%

    No Known Activations