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. 925
    Prev
    Next
    INDEX
    Explanations

    instances of the verb "talk" in various forms and contexts

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

    dialogue or discussions about talking

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

    talk about or to

    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
    LayoutConstraint
    -0.60
    Ỉ
    -0.57
    PageRoute
    -0.56
    frc
    -0.53
    IBarButtonItem
    -0.51
    DataAnnotations
    -0.51
    asley
    -0.51
    undi
    -0.51
    asteroide
    -0.51
    esModule
    -0.50
    POSITIVE LOGITS
     TALK
    1.05
    talk
    1.05
     talk
    1.03
    Talking
    1.02
     Talking
    0.98
     talked
    0.96
    talking
    0.95
     Talk
    0.95
     talking
    0.95
     Talks
    0.91
    Activations Density 0.023%

    No Known Activations