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-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 30320
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and their components in text

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

    mathematical expressions within parentheses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
    EndContext
    -0.30
     vain
    -0.28
     فريبيس
    -0.25
    Poloha
    -0.24
     annonces
    -0.24
    wachsene
    -0.23
    当
    -0.23
    床
    -0.23
    extAlignment
    -0.22
    RuleContext
    -0.22
    POSITIVE LOGITS
     ControllerBase
    0.63
    WebVitals
    0.61
    oredCriteria
    0.58
    dafx
    0.56
    haikusbot
    0.55
    ExtendWith
    0.55
     CreateTagHelper
    0.55
    factors
    0.55
    TargetException
    0.55
    yntaxException
    0.53
    Activations Density 0.732%

    No Known Activations