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. 8359
    Prev
    Next
    INDEX
    Explanations

    repeated phrases related to time or events occurring at specific moments

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

    "At" followed by a definite article ("the")

    np_acts-logits-general · gemini-2.0-flash

    at the

    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
     ddelweddau
    -0.77
    MemoryWarning
    -0.72
     незавершена
    -0.71
    oa̍t
    -0.71
     FetchType
    -0.70
     propOrder
    -0.66
    ="{{$
    -0.64
    lorette
    -0.63
    ImageContext
    -0.62
    Geplaatst
    -0.62
    POSITIVE LOGITS
     Likewise
    0.65
    SequentialGroup
    0.65
    tagHelperRunner
    0.65
     samtidig
    0.64
    tualmente
    0.61
    Likewise
    0.61
     likewise
    0.60
     meanwhile
    0.59
     apparente
    0.58
     Along
    0.57
    Activations Density 0.142%

    No Known Activations