© 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. 12276
    Prev
    Next
    INDEX
    Explanations

    references to characters and elements from the Star Wars franchise, particularly focusing on Darth Vader

    oai_token-act-pair · gpt-3.5-turbo

    references to iconic characters from the Star Wars franchise

    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
    uments
    -0.89
    ãĥ³ãĤ¸
    -0.76
    ome
    -0.75
    ledge
    -0.71
    ivities
    -0.71
    glers
    -0.70
    ablishment
    -0.69
    teen
    -0.68
    ¦
    -0.67
    ivity
    -0.67
    POSITIVE LOGITS
     Vader
    0.92
     Kenobi
    0.87
     Bane
    0.74
    ufact
    0.74
     horr
    0.72
     Hots
    0.70
    nces
    0.70
    Ton
    0.67
     Rack
    0.66
    halla
    0.66
    Activations Density 0.031%

    No Known Activations