© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 74254
    Prev
    Next
    INDEX
    Explanations

    mathematical symbols and expressions

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

    closing brackets or delimiters

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

    tokens that are parts of mathematical/LaTeX markup and related structural symbols (commands, backslashes, delimiters, and math-mode punctuation).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
    tagHelperRunner
    -0.73
    tvguidetime
    -0.66
     HasFactory
    -0.62
     SwitchCompat
    -0.60
     snippetHide
    -0.60
    matchCondition
    -0.60
     detailsLots
    -0.56
    bootstrapcdn
    -0.56
     BytesLike
    -0.55
     مرئيه
    -0.55
    POSITIVE LOGITS
    InjectAttribute
    0.56
    AccessorTable
    0.48
    DoubleQuotes
    0.42
     marbre
    0.38
     innamor
    0.38
     avoient
    0.36
     prestasi
    0.36
    UpInside
    0.36
     nôtre
    0.35
     inscrição
    0.35
    Activations Density 6.936%

    No Known Activations