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-2-2B
    4. Residual Stream - 65k
    5. 16-GEMMASCOPE-RES-65K
    6. 53504
    Prev
    Next
    INDEX
    Explanations

    numerical data and references related to engineering problems or studies

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

    technical jargon and citations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    XmlAccessType
    -0.52
     JAXBElement
    -0.48
    UnknownFields
    -0.48
    balleur
    -0.48
    ArgsConstructor
    -0.45
     departament
    -0.45
     prad
    -0.44
     Imperium
    -0.43
    Jereo
    -0.43
    EndInit
    -0.43
    POSITIVE LOGITS
    yntaxException
    0.64
     تضيفلها
    0.63
    principalColumn
    0.60
     ddelweddau
    0.60
    FormTagHelper
    0.60
     useParams
    0.57
    RegistryLite
    0.57
    UTELY
    0.55
    ftagPool
    0.55
     CURLOPT
    0.55
    Activations Density 0.290%

    No Known Activations