© 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. 2465
    Prev
    Next
    INDEX
    Explanations

    official titles and names of political figures

    oai_token-act-pair · gpt-3.5-turbo

    references to political figures and positions of authority

    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
     queen
    -0.80
     foss
    -0.76
     wheelchair
    -0.74
     destro
    -0.73
    emis
    -0.73
    sole
    -0.72
    elled
    -0.72
     stroke
    -0.72
    iscopal
    -0.71
    aukee
    -0.69
    POSITIVE LOGITS
     Candidate
    0.91
     Secretary
    0.90
     Agent
    0.89
     Clause
    0.87
     Box
    0.87
     Theory
    0.86
     Styles
    0.86
     Ambassador
    0.86
     Leader
    0.85
     Officer
    0.85
    Activations Density 0.118%

    No Known Activations