© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 3-RES-JB
    6. 23721
    Prev
    Next
    INDEX
    Explanations

    instances where there is juxtaposition or contrast between two ideas or concepts

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "although"

    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
    tnc
    -0.80
    hal
    -0.80
    lees
    -0.79
    enter
    -0.76
    elle
    -0.74
    Eye
    -0.72
    isible
    -0.72
    ath
    -0.71
    ais
    -0.69
    crop
    -0.69
    POSITIVE LOGITS
     acknowledging
    0.89
    chery
    0.81
    tons
    0.79
     retaining
    0.78
     agreeing
    0.78
     admittedly
    0.78
    yip
    0.77
     conced
    0.75
     browsing
    0.73
     technically
    0.72
    Activations Density 0.030%

    No Known Activations