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. 11-GEMMASCOPE-RES-16K
    6. 6754
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "first" and related phrases

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Hook Layer
    11
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     méri
    -0.94
    ukone
    -0.91
     Krieger
    -0.85
     Scrolls
    -0.84
    woordig
    -0.81
     Elli
    -0.80
     decorada
    -0.79
     Magdalene
    -0.79
    rapie
    -0.79
     Taras
    -0.78
    POSITIVE LOGITS
    FIRST
    1.60
     First
    1.58
     FIRST
    1.52
    First
    1.46
     first
    1.43
    first
    1.32
     first
    1.15
     getFirst
    1.14
     ersten
    1.07
    isFirst
    1.03
    Activations Density 0.123%

    No Known Activations