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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-2B
    7. Residual Stream - 65k
    9. 20-GEMMASCOPE-RES-65K
    11. 502
    Prev
    Next
    INDEX
    Explanations

    phrases indicating quality or comparison in various contexts

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

    evaluative language related to quality and qualifications

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

    equipped or qualifying

    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
     урна
    -0.56
    örande
    -0.54
    TestRunner
    -0.53
     Insee
    -0.51
    especie
    -0.48
     exprim
    -0.48
    InvalidProtocol
    -0.47
     abf
    -0.47
    needed
    -0.46
    angerschaft
    -0.46
    POSITIVE LOGITS
     equipped
    1.25
    Fully
    1.05
     Fully
    1.05
     fully
    1.04
     Equipped
    1.03
    equipped
    1.02
     armed
    0.97
     educated
    0.96
     adequately
    0.96
    Equipped
    0.95
    Activations Density 0.198%

    No Known Activations