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. 114546
    Prev
    Next
    INDEX
    Explanations

    the start of new sections or paragraphs in documents

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

    The token "bos" (beginning of sequence) marked between triple angle brackets, signaling the start of a new text section, document, or article.

    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
    GEBURTSDATUM
    -0.68
    makeConstraints
    -0.63
    addPreferredGap
    -0.63
    IntoConstraints
    -0.60
     queſta
    -0.60
    Vidite
    -0.60
     invokingState
    -0.57
    :✨
    -0.56
     disambiguazione
    -0.55
     المعيارى
    -0.55
    POSITIVE LOGITS
    onto
    0.43
     avail
    0.42
    flg
    0.42
    uff
    0.40
     available
    0.38
     Soler
    0.37
    rigan
    0.37
     aks
    0.36
    available
    0.35
    coste
    0.35
    Activations Density 0.057%

    No Known Activations