© 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. 11-RES-JB
    11. 24528
    Prev
    Next
    INDEX
    Explanations

    mentions of political figures and actions related to international conflicts, specifically involving Argentina and the UK

    oai_token-act-pair · gpt-3.5-turbo

    references to political events and figures, particularly relating to Argentina and the UK

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Hop
    -0.91
    Baltimore
    -0.91
     Hopkins
    -0.90
     Dalton
    -0.88
    soDeliveryDate
    -0.87
    Hunt
    -0.86
    Minnesota
    -0.82
    Han
    -0.77
     Huma
    -0.77
     Peel
    -0.76
    POSITIVE LOGITS
     Argentine
    2.75
     Argentina
    2.57
     Buenos
    2.56
     Argent
    2.52
     Aires
    1.93
     Falk
    1.85
     Fernandez
    1.64
     Chile
    1.59
     Paragu
    1.58
     Urug
    1.57
    Activations Density 0.160%

    No Known Activations