Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 11-RES-JB
    6. 19057
    Prev
    Next
    INDEX
    Explanations

    phrases denoting the passage of time

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "since."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    cipl
    -0.69
    abus
    -0.68
    vich
    -0.68
    igraph
    -0.67
    abulary
    -0.64
    iced
    -0.64
    itect
    -0.63
    erb
    -0.62
    ĺ
    -0.60
    okingly
    -0.60
    POSITIVE LOGITS
     then
    1.55
     THEN
    1.14
    rely
    1.10
    then
    1.05
    Then
    0.97
     Then
    0.96
     arriving
    0.90
     inception
    0.88
     2005
    0.80
     2009
    0.80
    Activations Density 0.067%

    No Known Activations