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. 18-GEMMASCOPE-RES-16K
    6. 983
    Prev
    Next
    INDEX
    Explanations

    requests for extensions of time in legal contexts

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

    durations of time

    np_acts-logits-general · gemini-2.0-flash

    time periods and durations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     HttpNotFound
    -0.89
    PerformLayout
    -0.85
     onCancelled
    -0.85
    complexContent
    -0.82
    mappedBy
    -0.78
    tvguidetime
    -0.77
    @[+][
    -0.76
    Hochspringen
    -0.72
    OCCURRED
    -0.70
     transfieras
    -0.70
    POSITIVE LOGITS
     time
    1.49
     Time
    1.21
    time
    1.13
    Time
    1.12
     TIME
    1.01
    时间
    1.00
     tiempo
    0.94
    時間
    0.89
     plazo
    0.89
    TIME
    0.89
    Activations Density 0.153%

    No Known Activations