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. 38755
    Prev
    Next
    INDEX
    Explanations

    terms related to railing systems and safety features

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
    ########.
    -0.50
    ⓧ
    -0.50
    HostException
    -0.50
    amethasone
    -0.50
    WebVitals
    -0.48
    GraphicsUnit
    -0.48
    itinéraire
    -0.47
    internalType
    -0.46
    كويكب
    -0.46
    يميديا
    -0.45
    POSITIVE LOGITS
     railings
    0.72
     railing
    0.65
     balustra
    0.62
     rails
    0.44
    rail
    0.42
     fence
    0.42
    السكان
    0.42
    ManagedBean
    0.41
    rails
    0.41
     rail
    0.41
    Activations Density 0.008%

    No Known Activations