© 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. 20565
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "Duke" with various endings and contexts

    oai_token-act-pair · gpt-3.5-turbo

    references to "Du" or similar terms related to specific entities, likely firms or institutions

    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
    ãģĨ
    -0.83
    giving
    -0.70
     Locked
    -0.70
     CPC
    -0.70
    )=(
    -0.70
    oulos
    -0.68
     Remastered
    -0.68
     Unsure
    -0.67
    nikov
    -0.66
    mberg
    -0.66
    POSITIVE LOGITS
    alog
    0.97
    pling
    0.96
    pee
    0.95
    alogy
    0.93
    isine
    0.90
    plet
    0.90
    cci
    0.89
    opoly
    0.87
    ples
    0.87
    plin
    0.86
    Activations Density 0.007%

    No Known Activations