© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 9748
    Prev
    Next
    INDEX
    Explanations

    corrections or revisions in texts

    oai_token-act-pair · gpt-3.5-turbo

    terms related to corrections and amendments in written documents

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    force
    -0.80
    uten
    -0.71
    irts
    -0.70
    wagen
    -0.66
    shape
    -0.65
     solvent
    -0.65
    usting
    -0.65
    itness
    -0.65
    rights
    -0.65
    powers
    -0.64
    POSITIVE LOGITS
    Correction
    1.00
    Correct
    0.87
     typo
    0.78
     cler
    0.72
     inaccurate
    0.70
     Timeline
    0.69
     erroneous
    0.67
     Clar
    0.66
     Emails
    0.65
    terday
    0.64
    Activations Density 0.025%

    No Known Activations