© 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. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5731
    Prev
    Next
    INDEX
    Explanations

    terms and technical language related to programming or computing environments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    760
    +0.10
    0.5%
    516
    +0.09
    0.4%
    235
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.10
    0.06
    158
    +0.09
    0.05
    284
    +0.09
    0.04
    Negative Logits
     Rankings
    -0.84
     Versus
    -0.80
     Fame
    -0.80
     Username
    -0.71
     territory
    -0.70
     Ranking
    -0.69
     Mile
    -0.67
    autions
    -0.66
     NX
    -0.66
     Seat
    -0.65
    POSITIVE LOGITS
    dylib
    1.00
    lib
    0.91
     lib
    0.87
    Lib
    0.83
    tur
    0.76
    abi
    0.75
    prim
    0.74
    common
    0.73
    imov
    0.71
    plugin
    0.70
    Activations Density 0.381%

    No Known Activations