Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 103095
    Prev
    Next
    INDEX
    Explanations

    code and configuration elements

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

    Variable names, attributes, and reference identifiers in technical code and documentation contexts.

    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
     cref
    -0.36
     steps
    -0.34
     متعلقه
    -0.32
     banda
    -0.32
    bootstrapcdn
    -0.31
    Autoritní
    -0.31
     
    -0.30
    WithMany
    -0.30
     Pingback
    -0.30
     pil
    -0.29
    POSITIVE LOGITS
     Paglinawan
    0.72
    SharedDtor
    0.70
    SourceChecksum
    0.68
    ंदीखरीदारी
    0.68
    pingente
    0.67
     queſta
    0.65
     kasarigan
    0.64
    SharedCtor
    0.60
    ágenes
    0.60
     mauva
    0.60
    Activations Density 0.055%

    No Known Activations