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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 14775
    Prev
    Next
    INDEX
    Explanations

    structured data and email headers in content

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

    Email headers (To, From, Subject, CC, BCC)

    np_acts-logits-general · gemini-2.0-flash

    email subject lines

    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
     Dietz
    -0.60
    webElementGuid
    -0.58
    RegistryLite
    -0.56
     onCreateView
    -0.55
     AdapterView
    -0.54
    mybatisplus
    -0.52
    vata
    -0.52
    SizeF
    -0.52
    CppMethod
    -0.51
    ContentAsync
    -0.51
    POSITIVE LOGITS
     Subject
    1.48
     subject
    1.46
    Subject
    1.43
    subject
    1.32
     SUBJECT
    1.23
    SUBJECT
    1.20
     subjects
    1.06
     subj
    1.04
     asunto
    0.92
     fubject
    0.91
    Activations Density 0.238%

    No Known Activations