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-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 30841
    Prev
    Next
    INDEX
    Explanations

    details related to pharmaceutical compositions and their preparation processes

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

    liquids and dosages

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
    rungsseite
    -0.52
     Paglinawan
    -0.47
    ConstraintMaker
    -0.44
     NSCoder
    -0.43
     poveznice
    -0.43
    KommentareTeilen
    -0.42
    Revenir
    -0.41
    SequentialGroup
    -0.41
    ⮕
    -0.40
    UnitTesting
    -0.40
    POSITIVE LOGITS
    rizona
    0.43
    Direct
    0.41
    رشف
    0.40
    ImageContext
    0.39
    MultipartFile
    0.38
    darah
    0.38
    లాలు
    0.38
     Scratch
    0.38
     Direct
    0.37
     liquid
    0.36
    Activations Density 0.066%

    No Known Activations