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-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 7364
    Prev
    Next
    INDEX
    Explanations

    information related to real estate and possibly client satisfaction.

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    concepts related to communication and education.

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

    phrases or concepts related to fundamental principles, core ideas, or essential truths. The neuron activates strongly for phrases like "education is the only way", "ignorance is bliss", "information is everything", and "communication is key", suggesting it detects statements presenting key insights or fundamental

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @bot
    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
     spacy
    -0.52
     ComVisible
    -0.42
     pensado
    -0.40
    enschappelijke
    -0.40
    โปร
    -0.39
     linkovi
    -0.39
    처럼
    -0.39
     voluto
    -0.38
    شوف
    -0.37
     pione
    -0.37
    POSITIVE LOGITS
     kunci
    0.85
     keys
    0.80
    
    0.80
     antidote
    0.74
     garantie
    0.73
    🔑
    0.73
    bewerken
    0.70
     clé
    0.70
     Keys
    0.70
     recipe
    0.69
    Activations Density 0.235%

    No Known Activations