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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 3-GEMMASCOPE-RES-16K
    6. 16119
    Prev
    Next
    INDEX
    Explanations

    occurrences of the substring "fr" in various contexts, likely indicating a focus on French linguistic elements or identifiers

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.3.hook_resid_post
    Hook Layer
    3
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    verwijspagina
    -0.82
     Schein
    -0.78
     Vickers
    -0.75
    Personendaten
    -0.71
     iſt
    -0.70
    })*/
    -0.69
    styleType
    -0.69
     Aitken
    -0.68
     Lockheed
    -0.66
    LEAN
    -0.65
    POSITIVE LOGITS
    fr
    2.74
     Fr
    2.62
     fr
    2.60
    Fr
    2.54
    FR
    2.41
     FR
    2.40
    fra
    1.45
     Fry
    1.44
    fri
    1.38
     fri
    1.28
    Activations Density 0.096%

    No Known Activations