© 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. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 5795
    Prev
    Next
    INDEX
    Explanations

    numerical data or values associated with measurements and statistical representations

    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
    27
    +0.10
    0.4%
    373
    +0.09
    0.4%
    204
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.10
    0.02
    432
    +0.09
    0.02
    156
    +0.09
    0.01
    Negative Logits
     outgoing
    -0.81
    Īè
    -0.72
    theless
    -0.72
     contro
    -0.68
    DEN
    -0.67
     curses
    -0.67
    TERN
    -0.66
    ij士
    -0.64
    VILLE
    -0.64
    HUD
    -0.62
    POSITIVE LOGITS
    amp
    1.39
    amps
    1.01
    lt
    0.88
     Amp
    0.79
    oise
    0.78
    oped
    0.77
    adium
    0.77
    amping
    0.75
    gt
    0.75
     amp
    0.73
    Activations Density 0.050%

    No Known Activations