© 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 EvalsExportsinterp-engineNEWAPI 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. 18476
    Prev
    Next
    INDEX
    Explanations

    words related to conflicts of interest, especially with a focus on CEOs and compensation

    oai_token-act-pair · gpt-3.5-turbo

    phrases or references related to starter packages or introductions

    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
     GOODMAN
    -0.79
     portions
    -0.69
     fractures
    -0.68
    ordable
    -0.68
    aukee
    -0.65
    etsk
    -0.64
     confounding
    -0.62
    byn
    -0.61
     impacts
    -0.61
    urally
    -0.61
    POSITIVE LOGITS
    ng
    0.71
    Redditor
    0.67
    erk
    0.67
    zie
    0.64
    Alpha
    0.63
     Phi
    0.63
    iw
    0.62
     UCHIJ
    0.62
     Plug
    0.62
    :[
    0.61
    Activations Density 0.000%

    No Known Activations