© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 1464
    Prev
    Next
    INDEX
    Explanations

    key elements and specifications in technical and product-related documents

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

    Programming code changes

    np_acts-logits-general · gemini-2.0-flash

    words after specific headings

    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
    TagMode
    -0.60
    AndEndTag
    -0.60
    GOTREF
    -0.58
    SequentialGroup
    -0.57
    IUrlHelper
    -0.57
    queryInterface
    -0.55
     виправивши
    -0.54
    httphttps
    -0.52
    Биография
    -0.50
     transfieras
    -0.49
    POSITIVE LOGITS
    ":[{
    0.41
    ❮
    0.39
     inoltre
    0.36
    RenderAtEndOf
    0.34
    ICR
    0.32
    qu
    0.31
     AllMovie
    0.31
    tagHelperRunner
    0.30
    equ
    0.30
     APR
    0.30
    Activations Density 0.758%

    No Known Activations