© 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. 1-RES-JB
    11. 12900
    Prev
    Next
    INDEX
    Explanations

    specific locations and geopolitical entities, particularly in the Balkans region

    oai_token-act-pair · gpt-3.5-turbo

    references to geopolitical regions and conflicts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    haw
    -0.80
    ulton
    -0.77
    hare
    -0.75
    HOU
    -0.73
    bery
    -0.71
    houses
    -0.71
    WARD
    -0.71
    uscript
    -0.70
    rams
    -0.70
    posts
    -0.70
    POSITIVE LOGITS
     Albania
    1.25
     Alban
    1.22
     Bosnia
    1.20
     Kosovo
    1.17
     Serbia
    1.16
     Yugoslavia
    1.11
     Monteneg
    1.00
     Airways
    0.95
     Herz
    0.94
     Serbian
    0.93
    Activations Density 0.011%

    No Known Activations