© 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-2-2B
    4. Residual Stream - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 651
    Prev
    Next
    INDEX
    Explanations

    references to book releases and their details

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

    references to media publications and the success of creative works

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

    released, published, coming out

    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-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
    webElementXpaths
    -0.61
    protoimpl
    -0.60
     CreateTagHelper
    -0.55
    BeginContext
    -0.54
     mariée
    -0.53
    RegistryLite
    -0.52
     ujednoznacz
    -0.51
     initComponents
    -0.49
    batore
    -0.49
    IconModule
    -0.49
    POSITIVE LOGITS
     released
    2.50
     release
    2.44
     releases
    2.17
     releasing
    2.11
     Released
    2.08
    released
    2.05
    Released
    2.01
    release
    1.97
     Release
    1.93
    Release
    1.85
    Activations Density 0.337%

    No Known Activations