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. 21334
    Prev
    Next
    INDEX
    Explanations

    phrases related to archival activities or official document storage

    oai_token-act-pair · gpt-3.5-turbo

    references to historical or organizational records and archives

    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
     pitch
    -0.77
     mesh
    -0.71
    wo
    -0.68
    weight
    -0.64
     displ
    -0.63
    killer
    -0.62
    model
    -0.62
     globally
    -0.62
    iazep
    -0.62
     stiff
    -0.61
    POSITIVE LOGITS
     Archives
    4.31
     archives
    2.45
     Archive
    2.29
     archive
    1.57
     Journals
    1.50
    archive
    1.48
    archives
    1.46
     Museum
    1.21
    Arch
    1.20
     Library
    1.19
    Activations Density 0.016%

    No Known Activations