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. 21017
    Prev
    Next
    INDEX
    Explanations

    financial or numerical information with emphasis on percentages

    oai_token-act-pair · gpt-3.5-turbo

    terms related to financial or statistical calculations, specifically those involving accounting for quantities or changes

    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
     slap
    -0.67
    pler
    -0.58
    mar
    -0.57
     Push
    -0.56
     Steps
    -0.54
     Volt
    -0.54
    uts
    -0.54
     suggestion
    -0.54
    lov
    -0.53
     Budapest
    -0.53
    POSITIVE LOGITS
     accounted
    3.79
     totaled
    1.62
     tallied
    1.51
     counted
    1.43
     accounts
    1.32
     accounting
    1.32
    account
    1.30
     account
    1.29
     constituted
    1.24
     averaged
    1.22
    Activations Density 0.012%

    No Known Activations