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. 709
    Prev
    Next
    INDEX
    Explanations

    text related to keeping or maintaining records, with a focus on criminal records

    oai_token-act-pair · gpt-3.5-turbo

    references to official records or documentation

    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
    cil
    -0.65
    bite
    -0.64
     abound
    -0.63
    ju
    -0.62
    ggle
    -0.61
    ractive
    -0.61
    flation
    -0.60
    atively
    -0.60
    aroo
    -0.59
    gent
    -0.58
    POSITIVE LOGITS
     records
    3.86
     Records
    2.77
     record
    2.63
     Record
    2.28
    record
    2.23
    Record
    1.84
     documents
    1.75
     recordings
    1.73
     logs
    1.53
     histories
    1.51
    Activations Density 0.009%

    No Known Activations