Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 816
    Prev
    Next
    INDEX
    Explanations

    modal verbs indicating potential or possibility in various contexts

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

    statements regarding the possibility and capability of deriving or seeing results in a mathematical or theoretical context

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

    potential states or transformations

    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
     kjem
    -0.55
     mijne
    -0.53
    providedIn
    -0.52
     semej
    -0.51
     Payable
    -0.51
     oportunidad
    -0.50
    AttributeSet
    -0.49
     payable
    -0.49
     sentire
    -0.49
     pagare
    -0.48
    POSITIVE LOGITS
    脚注の使い方
    0.80
     rewrite
    0.75
     rewritten
    0.70
    شهاد
    0.64
     XCTest
    0.61
     recast
    0.58
     rewriting
    0.58
    nemia
    0.58
    BeginContext
    0.56
    [])
    
    0.56
    Activations Density 0.230%

    No Known Activations