© 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. 5421
    Prev
    Next
    INDEX
    Explanations

    phrases describing methods or strategies

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

    Following "way" or "ways"

    np_acts-logits-general · gemini-2.0-flash

    way to or of

    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
    -1.09
    новниш
    -0.89
    TagMode
    -0.87
    Personensuche
    -0.87
     tartalomajánló
    -0.86
     pleaſure
    -0.82
     Roskov
    -0.82
     Meksiku
    -0.82
    CodedInputStream
    -0.80
     insuffisamment
    -0.80
    POSITIVE LOGITS
     ways
    0.90
     way
    0.87
     to
    0.70
     how
    0.68
    Ways
    0.65
     Ways
    0.63
     around
    0.58
     of
    0.57
     Way
    0.56
     WAYS
    0.55
    Activations Density 0.106%

    No Known Activations