© 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. 10-RES_POST_32K-OAI
    11. 29743
    Prev
    Next
    INDEX
    Explanations

    commands and instructions related to actions to take or subscriptions to choose

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    230
    +0.12
    0.6%
    345
    +0.12
    0.6%
    182
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.12
    -0.01
    130
    +0.12
    0.01
    125
    +0.09
    0.01
    Negative Logits
    imer
    -0.69
     disadvant
    -0.67
     Fatal
    -0.64
     encount
    -0.64
    iaries
    -0.62
     helic
    -0.62
     fibre
    -0.62
     skelet
    -0.62
    atic
    -0.59
     crawl
    -0.58
    POSITIVE LOGITS
    $$$$
    1.22
     Cancel
    0.76
     Sell
    0.70
    Button
    0.69
    currency
    0.68
    irmation
    0.67
    eem
    0.67
     Shipping
    0.66
    onz
    0.66
    Accept
    0.66
    Activations Density 0.025%

    No Known Activations