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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 8-RES-JB
    11. 7639
    Prev
    Next
    INDEX
    Explanations

    various instances of text being posted online on different platforms

    oai_token-act-pair · gpt-3.5-turbo

    instances of content being shared or posted online

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    rats
    -0.77
    TIT
    -0.67
    BILITIES
    -0.64
    ItemThumbnailImage
    -0.63
    amus
    -0.62
    uko
    -0.62
     ens
    -0.61
    displayText
    -0.61
    EH
    -0.61
    lement
    -0.60
    POSITIVE LOGITS
     behalf
    1.30
     Pastebin
    1.13
     Craigslist
    1.12
     Youtube
    1.07
     eBay
    1.06
     Facebook
    1.05
     YouTube
    1.05
     forums
    1.03
     Twitter
    1.02
     facebook
    0.99
    Activations Density 0.138%

    No Known Activations