© 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 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. 7118
    Prev
    Next
    INDEX
    Explanations

    conjunctions and phrases that connect ideas

    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
    393
    +0.16
    0.7%
    484
    +0.09
    0.4%
    561
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.16
    0.18
    0
    +0.09
    0.15
    484
    +0.09
    0.15
    Negative Logits
    acas
    -0.75
    ãĥ¼ãĥ
    -0.71
    actionDate
    -0.69
    uta
    -0.66
    uper
    -0.65
    ãĥ³ãĤ¸
    -0.64
    iste
    -0.64
    ulp
    -0.64
    urg
    -0.64
    owered
    -0.64
    POSITIVE LOGITS
     consequently
    1.57
     thus
    1.55
     hence
    1.55
     therefore
    1.53
     secondly
    1.47
     thereby
    1.36
     furthermore
    1.32
     moreover
    1.21
     then
    1.19
     vice
    1.11
    Activations Density 1.967%

    No Known Activations