© 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. 8-RES-JB
    11. 8371
    Prev
    Next
    INDEX
    Explanations

    phrases related to social media users and online interactions

    oai_token-act-pair · gpt-3.5-turbo

    references to user interactions on social media platforms

    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
     divest
    -0.73
     deregulation
    -0.70
    ierrez
    -0.70
     Ministers
    -0.68
     Presidents
    -0.67
    ospital
    -0.67
     funer
    -0.63
    AIDS
    -0.62
     executives
    -0.61
     dissolution
    -0.61
    POSITIVE LOGITS
    Redditor
    1.11
     commenter
    1.00
    username
    0.97
     reddit
    0.96
     Reddit
    0.94
     redd
    0.92
     Redd
    0.92
     Gamer
    0.89
     uploaded
    0.89
     upload
    0.88
    Activations Density 0.295%

    No Known Activations