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-2-2B
    4. Residual Stream - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 809
    Prev
    Next
    INDEX
    Explanations

    the presence of structured or formalized coding syntax

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

    programming structure and method definitions

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

    added # social

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
    はじめに
    -0.58
    mě
    -0.48
     donc
    -0.46
    yorsunuz
    -0.45
     وسلم
    -0.42
    ã
    -0.40
    ped
    -0.39
     "...
    -0.38
    комен
    -0.37
    ета
    -0.37
    POSITIVE LOGITS
    CloseOperation
    0.98
    MessageTagHelper
    0.96
     متعلقه
    0.92
    ArrowToggle
    0.80
     JpaRepository
    0.80
    expandindo
    0.79
    };*/
    0.79
    lapsingToolbar
    0.79
     '\\;'
    0.76
    ]--;
    0.76
    Activations Density 0.016%

    No Known Activations