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-IT
    4. Residual Stream - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 15951
    Prev
    Next
    INDEX
    Explanations

    references to quality management and continuous improvement methodologies

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

    Quality and process control

    np_acts-logits-general · gemini-2.0-flash

    quality control and improvement

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     coy
    -0.37
     dewy
    -0.37
    vog
    -0.36
     initComponents
    -0.35
    illow
    -0.35
    cstdio
    -0.34
    koff
    -0.34
     INDEPENDENT
    -0.33
     Breslau
    -0.32
    MathML
    -0.32
    POSITIVE LOGITS
     Lean
    0.70
    Lean
    0.68
     lean
    0.66
    WebElementEntity
    0.64
     Quality
    0.64
    Quality
    0.62
     Statistical
    0.59
    Statistical
    0.58
     ISO
    0.57
     defect
    0.57
    Activations Density 0.111%

    No Known Activations