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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    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. 5021
    Prev
    Next
    INDEX
    Explanations

    instances where something is described as being done perfectly or fitting perfectly

    oai_token-act-pair · gpt-3.5-turbo

    the word "perfectly" and its variations in various contexts

    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
     Sources
    -0.75
    edia
    -0.68
    Sources
    -0.65
    rollment
    -0.65
    ership
    -0.64
    onding
    -0.63
     STA
    -0.63
    ond
    -0.63
    jury
    -0.62
     Publications
    -0.62
    POSITIVE LOGITS
     perfectly
    3.41
     beautifully
    1.89
     wonderfully
    1.82
     neatly
    1.75
     perfect
    1.73
     brilliantly
    1.70
     nicely
    1.65
    perfect
    1.56
     seamlessly
    1.47
     correctly
    1.45
    Activations Density 0.009%

    No Known Activations