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-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 12393
    Prev
    Next
    INDEX
    Explanations

    specific ingredients and flavor descriptions related to culinary contexts

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

    Qualities of food or drink

    np_acts-logits-general · gemini-2.0-flash

    flavor and taste descriptors

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
     mourut
    -0.46
     doğum
    -0.44
     turística
    -0.43
     bezpieczeństwa
    -0.43
     Transplantation
    -0.43
     preventiva
    -0.42
     empresarial
    -0.42
     médicale
    -0.41
     pédagogique
    -0.41
     démocratique
    -0.41
    POSITIVE LOGITS
     flavor
    0.70
     taste
    0.63
     flavour
    0.59
     flavors
    0.58
    Autoritní
    0.57
     flavorful
    0.55
     informée
    0.55
     tastes
    0.54
    flavor
    0.51
    AndEndTag
    0.50
    Activations Density 0.073%

    No Known Activations