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. 3119
    Prev
    Next
    INDEX
    Explanations

    references to interest rates and financial terms related to loans

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

    Mentions "interest rate" or "judgment"

    np_acts-logits-general · gemini-2.0-flash

    judgment interest rates

    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
     nakalista
    -0.56
    ämä
    -0.49
    andro
    -0.49
     estekak
    -0.48
    νη
    -0.47
    tică
    -0.45
    issipp
    -0.43
     edi
    -0.43
    enzio
    -0.42
    scribe
    -0.41
    POSITIVE LOGITS
     interest
    1.30
    Interest
    1.22
    interest
    1.21
     Interest
    1.17
     compound
    0.97
     INTEREST
    0.95
    INTEREST
    0.93
     compounding
    0.93
     yield
    0.92
    compound
    0.91
    Activations Density 0.201%

    No Known Activations