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
    4. Residual Stream - 16k
    5. 21-GEMMASCOPE-RES-16K
    6. 7713
    Prev
    Next
    INDEX
    Explanations

    references to time and events in a chronological context

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

    references to a specific iteration or occasion in time—often contrasting with previous attempts or indicating plans to do it again or in the future.

    oai_token-act-pair · gpt-5Triggered by @arielgoldsteinlab
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    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
    ----</
    -0.37
     calendriers
    -0.32
    zenten
    -0.32
    cellona
    -0.29
     guste
    -0.28
     periodically
    -0.28
    现状
    -0.27
    хіво
    -0.27
    まず
    -0.26
    POSITIVE LOGITS
    今度は
    0.72
     diesmal
    0.66
    今度
    0.65
     again
    0.62
    again
    0.59
     tighter
    0.56
    第二次
    0.56
    二度
    0.55
    這次
    0.55
    Again
    0.54
    Activations Density 0.504%

    No Known Activations