© 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. 0-RES-JB
    11. 22746
    Prev
    Next
    INDEX
    Explanations

    references to a specific location named Winnipeg

    oai_token-act-pair · gpt-3.5-turbo

    mentions of "Winnipeg" and "Manitoba"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Davis
    -0.86
    Davis
    -0.75
     Cruz
    -0.75
    Sil
    -0.71
     DOJ
    -0.71
     Sect
    -0.70
    alde
    -0.69
     SR
    -0.67
     Scalia
    -0.66
    lyak
    -0.66
    POSITIVE LOGITS
     Winnipeg
    3.52
     Manitoba
    3.42
    ipeg
    2.76
     Manit
    1.95
     Saskatchewan
    1.68
    uckland
    1.53
     Sask
    1.51
     Tasmania
    1.42
     Moose
    1.41
     Auckland
    1.40
    Activations Density 0.020%

    No Known Activations