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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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 - 16k
    5. 25-GEMMASCOPE-RES-16K
    6. 3882
    Prev
    Next
    INDEX
    Explanations

    technical terms and methods related to programming

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 25-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_25/width_16k/average_l0_116
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     delivering
    -0.71
     developing
    -0.68
     promoting
    -0.65
     educating
    -0.65
     advancing
    -0.64
     performing
    -0.64
     helping
    -0.63
     producing
    -0.63
     managing
    -0.62
     registering
    -0.61
    POSITIVE LOGITS
     Adoption
    0.75
     Creation
    0.74
     Preparation
    0.73
     Detection
    0.73
     Deletion
    0.73
     Destruction
    0.72
     Submission
    0.72
     Removal
    0.72
     Extraction
    0.71
    Insertion
    0.71
    Activations Density 0.471%

    No Known Activations