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. 11153
    Prev
    Next
    INDEX
    Explanations

    technical terms related to engineering and mechanical systems.

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

    technical information related to design, engineering, and inventions.

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    concepts related to physical stability, balance, and motion control. This neuron appears to activate strongly for terms and phrases describing: 1. Stability and balance (e.g. "postural stability", "center of gravity") 2. Motion and movement (e.g. "flight

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @bot

    "Balance" or "stability"

    np_acts-logits-general · gemini-2.0-flash

    balance and stability

    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
    يكب
    -0.61
     throats
    -0.58
     Extragalactic
    -0.56
    mtrl
    -0.55
     fibrosis
    -0.50
     Laramie
    -0.49
    bitset
    -0.49
     hyal
    -0.49
    Хьажоргаш
    -0.49
     inauguration
    -0.48
    POSITIVE LOGITS
     balance
    1.88
     Balance
    1.62
    balance
    1.59
    Balance
    1.46
     balancing
    1.45
     BALANCE
    1.43
     stability
    1.42
     balanced
    1.41
     balances
    1.38
    balancing
    1.35
    Activations Density 0.258%

    No Known Activations