© 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. 0-RES-JB
    11. 1901
    Prev
    Next
    INDEX
    Explanations

    locations and names related to the Middle East, particularly focusing on Iraq and Syria

    oai_token-act-pair · gpt-3.5-turbo

    references to locations or entities associated with specific regions or groups

    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
     inf
    -0.73
     Qu
    -0.67
     cap
    -0.67
     Epic
    -0.67
     custom
    -0.65
     additive
    -0.64
     carbon
    -0.62
     Met
    -0.62
     items
    -0.61
     charg
    -0.61
    POSITIVE LOGITS
    iya
    4.52
    iyah
    2.17
    iy
    1.58
    ya
    1.53
    ia
    1.38
    yah
    1.37
    ija
    1.29
    yi
    1.28
    iae
    1.17
    bah
    1.15
    Activations Density 0.014%

    No Known Activations