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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 10131
    Prev
    Next
    INDEX
    Explanations

    phrases expressing language fluency and competence.

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

    phrases referring to being fluent in a language.

    oai_token-act-pair · gemini-1.5-flashTriggered by @bot

    mentions of languages and communication-related topics, especially in a professional context.

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    phrases related to speaking or language fluency.

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @bot

    Fluency or proficiency in a language or skill

    np_acts-logits-general · gemini-2.0-flash

    language proficiency

    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-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
    twimg
    -0.54
     TimeUnit
    -0.53
    φάλ
    -0.51
    ReusableCell
    -0.50
    Klem
    -0.50
     lccn
    -0.48
     '\\;'
    -0.48
     Hug
    -0.47
     facteurs
    -0.47
    фраструк
    -0.47
    POSITIVE LOGITS
     proficiency
    1.68
     mastered
    1.59
     proficient
    1.54
     mastery
    1.54
     mastering
    1.47
     skill
    1.39
     Proficiency
    1.37
     skills
    1.36
     Mastering
    1.31
     learned
    1.23
    Activations Density 0.374%

    No Known Activations