Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 17396
    Prev
    Next
    INDEX
    Explanations

    mentions or occurrences of the name "Koh"

    oai_token-act-pair · gpt-3.5-turbo

    mentions of specific individuals, particularly Harold Hongju Koh

    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
     daemon
    -0.73
    bot
    -0.69
    des
    -0.65
     compute
    -0.64
    atre
    -0.62
    ESA
    -0.61
    USE
    -0.60
     que
    -0.60
    loads
    -0.59
     con
    -0.58
    POSITIVE LOGITS
     Koh
    4.04
     Roh
    1.65
     Kah
    1.47
     Ko
    1.25
     Kahn
    1.25
     Kuh
    1.24
     Kau
    1.24
     Kitt
    1.24
     Boh
    1.24
     Koen
    1.18
    Activations Density 0.020%

    No Known Activations