Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 52
    Prev
    Next
    INDEX
    Explanations

    concepts related to problem-solving and improvement in various contexts

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

    Preceding "way" or "ways"

    np_acts-logits-general · gemini-2.0-flash

    ways and manners

    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
     كومونز
    -0.46
    Disliked
    -0.44
    когда
    -0.44
     pris
    -0.43
     Audiodateien
    -0.42
    τζ
    -0.42
    ăl
    -0.41
    Referenties
    -0.41
     Branch
    -0.41
     atsi
    -0.41
    POSITIVE LOGITS
     ways
    2.34
     way
    1.98
     Ways
    1.79
     WAYS
    1.64
    Ways
    1.62
    ways
    1.51
     maneras
    1.47
     maneiras
    1.47
     manières
    1.42
    way
    1.40
    Activations Density 0.315%

    No Known Activations