© 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
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 104139
    Prev
    Next
    INDEX
    Explanations

    mathematical symbols, notations, and references related to problems and theorems

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

    References to labeled elements (equations, lemmas, propositions, etc.) in mathematical/scientific papers, typically enclosed in \[ \] delimiters in LaTeX.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    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
     للمعارف
    -0.71
    ValueStyle
    -0.63
    enumi
    -0.62
    帖最后由
    -0.60
    enumii
    -0.58
    InjectAttribute
    -0.57
    dafx
    -0.55
    uxxxx
    -0.53
    脚注の使い方
    -0.53
    Personensuche
    -0.52
    POSITIVE LOGITS
    aughey
    0.40
    :✨
    0.36
    theoremstyle
    0.36
     мәкал
    0.33
    olol
    0.32
    Médaille
    0.31
    NotEqualTo
    0.31
    initState
    0.30
     })
    0.30
     cu
    0.29
    Activations Density 0.381%

    No Known Activations