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
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 53705
    Prev
    Next
    INDEX
    Explanations

    phrases related to disruptions or challenges in social interactions

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

    Tokens that appear in technical or conversational text where the content is being quoted, referenced, or continued from a previous thought. Often these tokens connect related parts of text, especially at the beginning or end of phrases.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     يتيمه
    -0.78
    rungsseite
    -0.56
    DockStyle
    -0.54
    __":
    
    -0.52
    })`
    -0.49
    "])
    
    -0.48
     esternos
    -0.48
    Portail
    -0.47
    expandindo
    -0.47
     CreateTagHelper
    -0.47
    POSITIVE LOGITS
    :✨
    0.39
    いいですね
    0.35
    のですね
    0.35
     cyber
    0.34
    жешь
    0.33
     unicórnio
    0.32
    んですね
    0.32
     asupra
    0.32
     eingeb
    0.32
    toMatchSnapshot
    0.32
    Activations Density 0.082%

    No Known Activations