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-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 902
    Prev
    Next
    INDEX
    Explanations

    occurrences of assertions and expectations within a test case context

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

    Tests or checks for code/properties

    np_acts-logits-general · gemini-2.0-flash

    testing code expects `return`

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 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
     ControllerBase
    -0.58
    Kaynakça
    -0.56
    WireFormatLite
    -0.51
     Chwiliwch
    -0.48
     الرياضيه
    -0.47
    Bakgrunnsstoff
    -0.47
     ſeveral
    -0.46
     ſtand
    -0.46
     ſta
    -0.45
     uſed
    -0.45
    POSITIVE LOGITS
    ngdoc
    0.49
     accurate
    0.41
     Works
    0.41
     when
    0.40
     basic
    0.40
     Worked
    0.38
    toBeTruthy
    0.38
    нулся
    0.38
     properties
    0.37
     mix
    0.37
    Activations Density 0.225%

    No Known Activations