© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    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. 402
    Prev
    Next
    INDEX
    Explanations

    concepts related to mechanisms and principles in scientific contexts

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

    Mentions "principle" or "mechanism"

    np_acts-logits-general · gemini-2.0-flash

    working principle and mechanism

    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
     AssemblyCulture
    -0.51
     Chwiliwch
    -0.50
     pinulongan
    -0.50
    RegressionTest
    -0.47
    .*")]
    -0.47
    BagConstraints
    -0.46
    esModule
    -0.45
    הערות
    -0.45
    queryInterface
    -0.44
    HomeAsUpEnabled
    -0.44
    POSITIVE LOGITS
    原理
    0.63
    princip
    0.44
    仕組み
    0.44
     mechan
    0.43
     mechanism
    0.42
    Princip
    0.42
     principle
    0.42
    mechan
    0.41
     principles
    0.41
     Mechan
    0.40
    Activations Density 0.910%

    No Known Activations