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. 20-GEMMASCOPE-RES-16K
    6. 1383
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing majority opinions or common experiences

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

    Preceding adverbs of certainty/frequency

    np_acts-logits-general · gemini-2.0-flash

    most followed by qualifiers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
    OGND
    -0.84
     AssemblyTitle
    -0.83
     وتسجيلات
    -0.72
     صوتيه
    -0.68
    HandlerContext
    -0.68
     tantôt
    -0.62
    quele
    -0.60
    parsedMessage
    -0.59
    يكب
    -0.59
     quelquefois
    -0.58
    POSITIVE LOGITS
     likely
    1.17
     importantly
    1.04
     of
    0.88
     notably
    0.81
    likely
    0.80
     Likely
    0.72
     assuredly
    0.71
     Importantly
    0.64
     everyone
    0.62
     often
    0.61
    Activations Density 0.070%

    No Known Activations