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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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. 107052
    Prev
    Next
    INDEX
    Explanations

    references to programming concepts, particularly related to code migration and execution

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

    programming and code

    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
     town
    -0.28
    1
    -0.27
     posesión
    -0.27
    antas
    -0.26
     froide
    -0.26
    wachs
    -0.25
    2
    -0.25
    line
    -0.25
    8
    -0.24
    BeforeClass
    -0.24
    POSITIVE LOGITS
     незавершена
    0.93
     виправивши
    0.70
     Spoiler
    0.68
     يتيمه
    0.66
    ſelves
    0.66
     betweenstory
    0.66
     Signalez
    0.65
    StructEnd
    0.65
     endblock
    0.65
    __':
    
    0.64
    Activations Density 0.134%

    No Known Activations