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. 122757
    Prev
    Next
    INDEX
    Explanations

    technical specifications and references related to semiconductor devices

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

    Special characters and tokens in technical contexts, especially in code, mathematical notation, citations, and URIs.

    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
    انجليز
    -0.52
     Valdés
    -0.40
    XmlAccessType
    -0.39
    postmedia
    -0.35
    setcounter
    -0.35
     Domínguez
    -0.34
    HasColumnName
    -0.34
     kebenaran
    -0.33
    ünstler
    -0.31
    >{@
    -0.31
    POSITIVE LOGITS
     ब्रेकडाउन
    0.55
    WireFormatLite
    0.52
    AndEndTag
    0.51
     surla
    0.49
    Tembelea
    0.48
    न्दर्भ
    0.48
    mybatisplus
    0.48
    WriteTagHelper
    0.48
     uLocal
    0.48
    LEncoder
    0.48
    Activations Density 0.516%

    No Known Activations