© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-2B
    7. Residual Stream - 16k
    9. 16-GEMMASCOPE-RES-16K
    11. 13623
    Prev
    Next
    INDEX
    Explanations

    instances of the word "and" and phrases that connect multiple ideas or conditions

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

    followed by gerunds

    np_acts-logits-general · gemini-2.0-flash

    and/or followed by words ending in -ing

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     démocr
    -0.69
     sağlar
    -0.65
     modulated
    -0.62
     gangen
    -0.61
     disediakan
    -0.61
     kullanılır
    -0.61
     fringed
    -0.61
    RTHOOK
    -0.60
     ditetapkan
    -0.59
    bewah
    -0.58
    POSITIVE LOGITS
     being
    1.66
     having
    1.56
     getting
    1.54
     making
    1.48
     taking
    1.39
     giving
    1.36
    being
    1.34
     finding
    1.26
     doing
    1.26
     providing
    1.24
    Activations Density 0.239%

    No Known Activations