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. 966
    Prev
    Next
    INDEX
    Explanations

    web-related terms or phrases

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "Web"

    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
    ãĢİ
    -0.79
    â̦]
    -0.74
    â̦..
    -0.73
    ãĢı
    -0.72
    â̦â̦
    -0.69
     republican
    -0.66
     whilst
    -0.66
     mould
    -0.65
    eli
    -0.65
     colours
    -0.64
    POSITIVE LOGITS
     Web
    3.69
    Web
    2.55
     web
    2.39
    web
    2.33
     Webs
    2.11
     Internet
    1.88
     webs
    1.70
     webpage
    1.57
     Website
    1.53
     websites
    1.52
    Activations Density 0.008%

    No Known Activations