© 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. 19003
    Prev
    Next
    INDEX
    Explanations

    phrases related to specific technical terms or acronyms, particularly "DK" and "MAC."

    oai_token-act-pair · gpt-3.5-turbo

    acronyms and shorthand that refer to organizations or technical terms

    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
     Catalan
    -0.86
     Catalonia
    -0.73
    Copyright
    -0.69
    Catal
    -0.69
    upiter
    -0.69
     Telescope
    -0.69
     Plato
    -0.69
     bladder
    -0.67
     Moses
    -0.66
     Psal
    -0.66
    POSITIVE LOGITS
     IG
    3.69
     DK
    1.72
     MAC
    1.45
     AH
    1.36
    CG
    1.27
     CG
    1.26
     MU
    1.25
     GA
    1.20
     GH
    1.15
    MAC
    1.14
    Activations Density 0.062%

    No Known Activations