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. 3076
    Prev
    Next
    INDEX
    Explanations

    phrases related to technology and the internet, specifically mentioning "World Wide" or "Wide."

    oai_token-act-pair · gpt-3.5-turbo

    references to the "Wide" network or system, as well as terms related to offensive and defensive strategies

    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
     pret
    -0.74
     lev
    -0.74
     delinqu
    -0.73
     salvage
    -0.73
     summons
    -0.72
     spirit
    -0.71
     val
    -0.71
     martyr
    -0.69
     impression
    -0.69
     plac
    -0.67
    POSITIVE LOGITS
     Wide
    3.48
    Wide
    2.57
     Tight
    1.47
     Offensive
    1.35
     Defensive
    1.27
     Flavoring
    1.14
     Milky
    1.12
     Large
    1.12
     Telesc
    1.07
     Dumb
    1.05
    Activations Density 0.027%

    No Known Activations