© 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
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 3-RES-JB
    11. 9011
    Prev
    Next
    INDEX
    Explanations

    comparisons between different entities or concepts

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating differences or distinctions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    wash
    -0.66
     fueled
    -0.66
    escape
    -0.65
     contaminated
    -0.65
    bye
    -0.64
     simulator
    -0.63
    don
    -0.60
     crawled
    -0.60
    ocalypse
    -0.59
    ded
    -0.59
    POSITIVE LOGITS
     sexes
    0.80
    lihood
    0.76
     different
    0.76
     chronological
    0.75
    ospels
    0.73
    asury
    0.73
    inx
    0.73
     differing
    0.73
     disparate
    0.70
     genders
    0.70
    Activations Density 0.189%

    No Known Activations