© 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. 3-RES-JB
    11. 24169
    Prev
    Next
    INDEX
    Explanations

    mentions of geographical locations related to the Gulf

    oai_token-act-pair · gpt-3.5-turbo

    references to the Gulf region

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ATURE
    -0.74
    66666666
    -0.74
    IENCE
    -0.70
    nings
    -0.70
    onsense
    -0.69
    enance
    -0.66
    lihood
    -0.65
    semble
    -0.64
    byter
    -0.63
    BuyableInstoreAndOnline
    -0.62
    POSITIVE LOGITS
    stream
    0.94
     Gulf
    0.89
     Coast
    0.88
     Cooperation
    0.85
     monarch
    0.76
     Airways
    0.75
    raits
    0.74
     coasts
    0.73
     Kuwait
    0.71
     Islands
    0.71
    Activations Density 0.005%

    No Known Activations