© 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. 24369
    Prev
    Next
    INDEX
    Explanations

    web links or handles on social media platforms

    oai_token-act-pair · gpt-3.5-turbo

    keywords related to technology or digital platforms, particularly URLs and social media handles

    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
    pread
    -0.71
    iola
    -0.69
     baggage
    -0.68
    ilda
    -0.67
    rig
    -0.63
    otiation
    -0.63
    ogue
    -0.61
    wich
    -0.61
     haul
    -0.60
    ingle
    -0.59
    POSITIVE LOGITS
    ··
    0.72
    é¾įåĸļ士
    0.72
     Delete
    0.70
    ©¶æ
    0.69
    cms
    0.68
     Accessed
    0.65
     ancest
    0.65
     Tactics
    0.65
     Croat
    0.62
     Recon
    0.62
    Activations Density 0.313%

    No Known Activations