© 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. 4-RES-JB
    11. 20574
    Prev
    Next
    INDEX
    Explanations

    names of people with the prefix "Ed" followed by strong emotions or actions

    oai_token-act-pair · gpt-3.5-turbo

    names, specifically those that start with "Ed" or are variations of that name

    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
    natureconservancy
    -0.76
    icipated
    -0.73
    nown
    -0.72
    accompan
    -0.70
    £ı
    -0.70
    kat
    -0.66
    ngth
    -0.66
     Cosponsors
    -0.66
    ¬¼
    -0.65
    heses
    -0.64
    POSITIVE LOGITS
    abase
    0.94
    ktop
    0.70
    haus
    0.70
    hack
    0.65
     Garland
    0.65
     omission
    0.64
     Vers
    0.62
     Animation
    0.62
     Louie
    0.61
    geist
    0.61
    Activations Density 0.231%

    No Known Activations