© 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. 0-RES-JB
    11. 16824
    Prev
    Next
    INDEX
    Explanations

    phrases beginning with 'Do'

    oai_token-act-pair · gpt-3.5-turbo

    the imperative form of the verb "do."

    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
    alus
    -0.73
    inement
    -0.64
    )=(
    -0.63
    aced
    -0.60
    ictions
    -0.59
     fixture
    -0.59
    ylum
    -0.58
    abl
    -0.57
    etter
    -0.56
    icket
    -0.56
    POSITIVE LOGITS
     Do
    3.23
    Do
    2.27
     DO
    1.74
    do
    1.72
     Does
    1.71
     Did
    1.65
    DO
    1.60
     Are
    1.51
     Have
    1.35
     do
    1.32
    Activations Density 0.008%

    No Known Activations