© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 7378
    Prev
    Next
    INDEX
    Explanations

    references to specific years and dates

    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
    512
    +0.09
    0.4%
    241
    +0.09
    0.4%
    739
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.09
    0.04
    306
    +0.09
    0.04
    430
    +0.09
    0.04
    Negative Logits
    avorite
    -0.81
     undermin
    -0.79
    ruciating
    -0.75
    ITH
    -0.74
    iasco
    -0.73
    unal
    -0.70
    SPONSORED
    -0.67
    arranted
    -0.67
    efficients
    -0.67
    FU
    -0.67
    POSITIVE LOGITS
     onwards
    0.82
     Publisher
    0.71
     onward
    0.70
     reprinted
    0.70
     release
    0.66
     Collection
    0.66
     Final
    0.65
     Release
    0.65
     EDITION
    0.64
     Advent
    0.62
    Activations Density 0.193%

    No Known Activations