© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    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. 7069
    Prev
    Next
    INDEX
    Explanations

    connections to economic and resource-based themes

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

    preceding "rather" or "instead"

    np_acts-logits-general · gemini-2.0-flash

    instead or rather

    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
    lapsingToolbar
    -0.65
     Infórmanos
    -0.64
    SupportActionBar
    -0.61
    ++
    
    -0.60
    ToListAsync
    -0.58
    portál
    -0.57
    monials
    -0.56
    thâu
    -0.55
     Courant
    -0.55
     Públicas
    -0.55
    POSITIVE LOGITS
     instead
    4.13
     rather
    4.02
    instead
    3.79
    rather
    3.55
    Instead
    3.44
     Instead
    3.40
     Rather
    3.35
    Rather
    3.33
     anstatt
    3.07
     plutôt
    3.03
    Activations Density 2.131%

    No Known Activations