© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 5065
    Prev
    Next
    INDEX
    Explanations

    keywords related to decks, potentially in the context of card games or construction projects

    oai_token-act-pair · gpt-3.5-turbo

    mentions of card decks in a game context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    hots
    -0.74
    outheastern
    -0.71
    orically
    -0.70
    untary
    -0.68
    Ø©
    -0.68
    ê
    -0.68
    ache
    -0.67
     Nort
    -0.65
     voluntary
    -0.64
    IED
    -0.63
    POSITIVE LOGITS
     deck
    3.87
     decks
    3.35
     Deck
    2.91
    deck
    2.73
     cards
    1.42
     card
    1.41
     archetype
    1.26
     Cards
    1.24
     board
    1.23
     Hearthstone
    1.21
    Activations Density 0.024%

    No Known Activations