© 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. 4-RES-JB
    11. 14227
    Prev
    Next
    INDEX
    Explanations

    references to the country "Mali."

    oai_token-act-pair · gpt-3.5-turbo

    references to countries, particularly in Africa

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

    No Comments

    Negative Logits
    itbart
    -0.80
     accur
    -0.74
    ellation
    -0.73
    rawdownloadcloneembedreportprint
    -0.70
    essler
    -0.70
    initions
    -0.67
    viation
    -0.67
    Madison
    -0.67
    enegger
    -0.65
    alty
    -0.65
    POSITIVE LOGITS
     Niger
    1.03
     Senegal
    1.00
     Laos
    1.00
     Cameroon
    0.97
     Congo
    0.97
     Nigeria
    0.97
     Angola
    0.88
     Erit
    0.87
     Africa
    0.87
     Ghana
    0.87
    Activations Density 0.025%

    No Known Activations