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 - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 16150
    Prev
    Next
    INDEX
    Explanations

    statements related to sports events and their outcomes

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

    preceding attribution or quotes

    np_acts-logits-general · gemini-2.0-flash

    attributing statements from people

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     nôtre
    -0.49
     aislada
    -0.47
     ferons
    -0.47
     masculino
    -0.46
     choisissez
    -0.45
     kommunen
    -0.45
     différente
    -0.44
     Económica
    -0.44
     adecuado
    -0.43
     sebagainya
    -0.43
    POSITIVE LOGITS
     former
    1.40
     veteran
    1.21
    former
    1.16
     fellow
    1.13
     longtime
    1.13
     Former
    1.10
     actress
    1.08
     comedian
    1.03
    Former
    1.03
     actor
    1.03
    Activations Density 0.234%

    No Known Activations