© 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-IT
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 13594
    Prev
    Next
    INDEX
    Explanations

    phrases indicating composition and structure

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

    preceding "of" that describe a composition

    np_acts-logits-general · gemini-2.0-flash

    composed of / consists of / comprised of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    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
    posedge
    -0.59
     שוליים
    -0.56
     виправивши
    -0.54
    原始内容存档于
    -0.52
    verwijspagina
    -0.52
    Zeneca
    -0.50
    AnimationsModule
    -0.49
    <bos>
    -0.48
     errno
    -0.48
    GIF
    -0.47
    POSITIVE LOGITS
     consists
    1.11
     consisted
    1.08
     comprised
    1.07
     composed
    1.01
     consisting
    1.00
     comprises
    0.98
     comprise
    0.93
     consist
    0.91
     comprising
    0.85
     består
    0.84
    Activations Density 0.286%

    No Known Activations