© 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. 0-RES-JB
    11. 14801
    Prev
    Next
    INDEX
    Explanations

    words related to administrative tasks and processes

    oai_token-act-pair · gpt-3.5-turbo

    references to administrative structures or actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Venom
    -0.78
    brace
    -0.72
     Shape
    -0.72
    eared
    -0.72
    aired
    -0.71
    Pred
    -0.67
     Cage
    -0.67
     Scythe
    -0.66
    ipples
    -0.66
    Shape
    -0.66
    POSITIVE LOGITS
     administrative
    3.55
     Administrative
    2.60
     administr
    2.18
    administ
    2.12
    Administ
    1.85
     administ
    1.61
     bureaucratic
    1.61
     administrator
    1.57
     admin
    1.56
     administrators
    1.55
    Activations Density 0.017%

    No Known Activations