© 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. 4-RES-JB
    11. 1393
    Prev
    Next
    INDEX
    Explanations

    technical descriptions of products

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to product features and design

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

    No Comments

    Negative Logits
    udo
    -0.94
     Presbyter
    -0.88
    pee
    -0.86
     pse
    -0.85
    iatus
    -0.81
    tarian
    -0.80
    ilk
    -0.80
    umably
    -0.79
     externalToEVAOnly
    -0.77
    odox
    -0.77
    POSITIVE LOGITS
     thrilling
    1.01
     sleek
    0.93
     stylish
    0.92
     thrill
    0.91
     unbeat
    0.90
     innovative
    0.90
     luxurious
    0.87
     immersive
    0.86
    eaturing
    0.86
     gorgeous
    0.86
    Activations Density 1.282%

    No Known Activations