© 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. 1-RES-JB
    11. 22374
    Prev
    Next
    INDEX
    Explanations

    mentions of royalty and royal settings

    oai_token-act-pair · gpt-3.5-turbo

    references to thrones and palaces, indicating themes of power and nobility

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Morse
    -0.72
    href
    -0.71
    ATIONAL
    -0.70
    went
    -0.69
    ãģ¦
    -0.67
     Greenberg
    -0.66
     Cooperative
    -0.66
    aturdays
    -0.66
    é¾
    -0.63
    KY
    -0.63
    POSITIVE LOGITS
     throne
    1.34
     Throne
    0.91
     palace
    0.88
    holders
    0.87
    holder
    0.85
     coffers
    0.84
    ilion
    0.83
    pole
    0.82
     crowned
    0.82
     flush
    0.80
    Activations Density 0.004%

    No Known Activations