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. 62188
    Prev
    Next
    INDEX
    Explanations

    specific food items or ingredients

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

    Common connecting words, punctuation marks, or phrase completions that link content together or finalize a thought in diverse textual contexts.

    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
     Photocase
    -0.51
     arşivlendi
    -0.47
     EconPapers
    -0.43
     nonUne
    -0.41
    knex
    -0.39
    횟
    -0.39
     périodes
    -0.38
     noticia
    -0.37
     pierna
    -0.36
     cifras
    -0.36
    POSITIVE LOGITS
    ACHUSET
    0.48
     Syndicate
    0.44
    uros
    0.43
    logic
    0.42
    AutoScaleMode
    0.41
    rad
    0.41
    commod
    0.40
    bnf
    0.40
    lish
    0.40
    nesium
    0.40
    Activations Density 0.059%

    No Known Activations