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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 21245
    Prev
    Next
    INDEX
    Explanations

    mentions of sports players and their activities

    oai_token-act-pair · gpt-3.5-turbo

    references to "players" in various contexts within sports or gaming

    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
     pale
    -0.73
     acid
    -0.71
     darn
    -0.69
     transit
    -0.68
     gas
    -0.67
     tube
    -0.66
     scept
    -0.65
     gust
    -0.64
     alarm
    -0.64
     ass
    -0.64
    POSITIVE LOGITS
     Players
    3.67
    Players
    2.55
     Player
    1.97
    players
    1.75
     Playing
    1.52
     Fans
    1.45
    Player
    1.43
    player
    1.35
     players
    1.33
     Characters
    1.33
    Activations Density 0.023%

    No Known Activations