© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 3-RES-JB
    11. 5453
    Prev
    Next
    INDEX
    Explanations

    words that are preludes or introductions to a description or excerpt

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "Desc" and variations of it, indicating descriptions or excerpts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    tery
    -0.68
    lift
    -0.67
     Wonderland
    -0.66
    flix
    -0.66
     heaven
    -0.64
     Bir
    -0.62
     whip
    -0.62
    strap
    -0.61
    heed
    -0.61
     loose
    -0.60
    POSITIVE LOGITS
    ript
    1.45
    ribing
    1.32
    ribed
    1.24
    ribe
    1.15
    ribes
    1.06
    endants
    1.06
    desc
    1.01
    ended
    0.98
    endant
    0.96
     Desc
    0.94
    Activations Density 0.005%

    No Known Activations