© 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
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 5892
    Prev
    Next
    INDEX
    Explanations

    code-related structures and elements, particularly in programming syntax and logic

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Geiſt
    -0.59
     zwiſchen
    -0.59
     CreateTagHelper
    -0.58
    HtmlAttribute
    -0.57
    tagHelperRunner
    -0.56
    saraba
    -0.56
     bénévol
    -0.56
    󠀠
    -0.56
     sostenibilidad
    -0.55
    setVerticalGroup
    -0.55
    POSITIVE LOGITS
     path
    0.59
    path
    0.59
     pathway
    0.52
    Path
    0.52
     Path
    0.51
    breadcrumb
    0.45
     navig
    0.42
    PATH
    0.42
     journey
    0.41
    辿
    0.41
    Activations Density 0.809%

    No Known Activations