© 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. 0-RES-JB
    11. 10220
    Prev
    Next
    INDEX
    Explanations

    instances of commands or instructions being given

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "ordered" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    develop
    -0.69
    ãĤ¤ãĥĪ
    -0.68
    PsyNetMessage
    -0.64
    Camp
    -0.60
    AGE
    -0.59
     exploit
    -0.59
     Lur
    -0.57
    riv
    -0.57
    uary
    -0.57
    adem
    -0.56
    POSITIVE LOGITS
     ordered
    3.70
     ordering
    2.29
    ordered
    2.16
     ORDER
    1.92
     orders
    1.91
     order
    1.88
     Orders
    1.72
    Order
    1.70
     Order
    1.66
    order
    1.61
    Activations Density 0.011%

    No Known Activations