© 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. 4-RES-JB
    11. 7413
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a sequence of actions or events

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "before" and related phrases indicating prior actions or events

    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
     Cosponsors
    -0.70
    ARI
    -0.70
    arios
    -0.67
    ARP
    -0.65
     Bord
    -0.64
     Preservation
    -0.64
    WER
    -0.62
     unfocusedRange
    -0.62
     Canaver
    -0.62
    ãĤ´
    -0.62
    POSITIVE LOGITS
    joining
    0.78
     embark
    0.70
    isode
    0.68
    asha
    0.68
    asus
    0.67
     expires
    0.65
     succ
    0.65
     ceases
    0.65
     arrive
    0.64
    ulent
    0.64
    Activations Density 0.167%

    No Known Activations