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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 1010
    Prev
    Next
    INDEX
    Explanations

    programming language elements related to binary expressions and boolean operations

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

    Code, operations, and expressions

    np_acts-logits-general · gemini-2.0-flash

    binary arithmetic operations

    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
    ThemeOverlay
    -0.70
    featureID
    -0.68
    ftagPool
    -0.66
    DockStyle
    -0.65
    richTextPanel
    -0.65
    MLLoader
    -0.64
    GraphicsUnit
    -0.63
    ]")]
    -0.63
    期刊论文
    -0.59
    providedIn
    -0.58
    POSITIVE LOGITS
    잃
    0.37
    verbs
    0.33
     verb
    0.32
     verbs
    0.32
    SafeMath
    0.30
     hook
    0.29
     înd
    0.29
    Arithmetic
    0.29
    خانه
    0.28
     gegenüber
    0.28
    Activations Density 0.499%

    No Known Activations