© 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. 2-RES-JB
    11. 13759
    Prev
    Next
    INDEX
    Explanations

    information or requests related to subscribing, donations, and newsletters

    oai_token-act-pair · gpt-3.5-turbo

    references to signing up for newsletters and online subscriptions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     abolished
    -0.52
    ridden
    -0.51
     centrif
    -0.49
     eroded
    -0.49
    abal
    -0.49
     marginal
    -0.49
    Qaida
    -0.48
     interrogated
    -0.48
    ['
    -0.48
    induced
    -0.48
    POSITIVE LOGITS
     Submit
    0.68
     THANK
    0.65
     Patreon
    0.64
     VIDEOS
    0.62
     SERVICES
    0.62
     reddit
    0.60
     :)
    0.60
     Email
    0.60
     appre
    0.60
     YOUR
    0.59
    Activations Density 4.375%

    No Known Activations