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. 130950
    Prev
    Next
    INDEX
    Explanations

    terms related to commitment and personal actions or decisions

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

    Verbs or action words that indicate intention, commitment, purpose, or agency within statements describing human behavior or decision-making.

    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
    intios
    -0.41
     Eller
    -0.41
    眼前
    -0.40
     gặp
    -0.39
    醐
    -0.36
    ệm
    -0.34
     rencontré
    -0.34
     incontra
    -0.34
    layoutControl
    -0.33
     CardView
    -0.33
    POSITIVE LOGITS
    WriteTagHelper
    0.71
    RenderAtEndOf
    0.64
    ########.
    0.62
     surla
    0.61
     AssemblyCulture
    0.59
    TargetException
    0.58
     Administrativna
    0.55
    ftagPool
    0.50
    tagHelperRunner
    0.49
     délib
    0.49
    Activations Density 0.254%

    No Known Activations