Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 20-GEMMASCOPE-RES-16K
    6. 13695
    Prev
    Next
    INDEX
    Explanations

    references to quality, accuracy, and performance in various contexts

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

    quality or correctness

    np_acts-logits-general · gemini-2.0-flash

    accuracy and performance evaluation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
     disambiguazione
    -0.62
    ंदीखरीदारी
    -0.62
     الحره
    -0.62
    Хьажоргаш
    -0.61
     الرياضيه
    -0.61
     betweenstory
    -0.61
    دانشنامهٔ
    -0.59
    AndEndTag
    -0.59
    SourceChecksum
    -0.57
    IntoConstraints
    -0.57
    POSITIVE LOGITS
     efficacy
    0.38
     berkualitas
    0.37
     efficacité
    0.36
    efficacité
    0.34
     rodillas
    0.33
     precisión
    0.32
     consideración
    0.32
     pneus
    0.32
    usercontent
    0.31
     dikatakan
    0.31
    Activations Density 0.369%

    No Known Activations