© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 3716
    Prev
    Next
    INDEX
    Explanations

    references to time-related expressions and actions

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

    phrases containing the word "time" in contemplative or decisive contexts, often suggesting a moment of reflection or action.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    phrases signaling that the appropriate moment has arrived to begin or do something (the “it’s time to/for …” call-to-action construction).

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the phrase "it's time to" or variations suggesting that the moment has arrived to take action.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    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
     Zust
    -0.36
     únicamente
    -0.36
     strictement
    -0.35
     Asamblea
    -0.34
     varied
    -0.33
     olmak
    -0.33
     mármol
    -0.32
     grabación
    -0.31
     zostanie
    -0.31
     camioneta
    -0.31
    POSITIVE LOGITS
    そろそろ
    0.98
    Tikang
    0.78
    writeField
    0.70
    finally
    0.67
     '\\;'
    0.66
    useNewUrlParser
    0.65
     наконец
    0.65
    VersionUID
    0.65
    出版年
    0.64
     finally
    0.63
    Activations Density 0.112%

    No Known Activations