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

    Neuronpedia

    Jacobian LensNEW
    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. 10-RES-JB
    11. 5902
    Prev
    Next
    INDEX
    Explanations

    social media-related content such as posts, online comments, and articles, especially focusing on those that may be provocative or sensitive in nature

    oai_token-act-pair · gpt-3.5-turbo

    references to online postings and social media content

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Quart
    -0.85
    ayne
    -0.74
     endurance
    -0.74
     strugg
    -0.71
     lineback
    -0.64
    Ĭ±
    -0.64
     chromos
    -0.63
     Chak
    -0.63
    igure
    -0.63
     negotiators
    -0.63
    POSITIVE LOGITS
     anonymously
    1.07
     online
    1.05
     spam
    1.02
     uploaded
    1.00
     inciting
    0.97
     containing
    0.96
     depicting
    0.94
     copyrighted
    0.93
    vertisements
    0.93
    Reddit
    0.92
    Activations Density 0.276%

    No Known Activations