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-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 15496
    Prev
    Next
    INDEX
    Explanations

    cryptographic or computational data elements, such as hexadecimal values or representations

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

    Hexadecimal strings or UUIDs

    np_acts-logits-general · gemini-2.0-flash

    hexadecimal strings

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
    DoubleQuotes
    -0.63
     CreateTagHelper
    -0.57
    himovic
    -0.57
    LookAnd
    -0.57
     Signalez
    -0.56
    LoadScene
    -0.56
    RegressionTest
    -0.54
    ITHER
    -0.53
    $_['
    -0.52
     Bourgoin
    -0.52
    POSITIVE LOGITS
     Shakspeare
    0.58
     [](
    0.57
     fleste
    0.56
    mybatisplus
    0.54
     detached
    0.54
     quæ
    0.54
    ніципалі
    0.52
     Antiquity
    0.52
     kenzo
    0.51
     življen
    0.51
    Activations Density 0.143%

    No Known Activations