© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 32703
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the concept of time, particularly focusing on duration and past events

    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
    359
    +0.11
    0.5%
    214
    +0.11
    0.5%
    47
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    136
    +0.11
    0.04
    171
    +0.11
    0.03
    458
    +0.10
    0.04
    Negative Logits
     Gund
    -0.69
     BMC
    -0.64
     FTA
    -0.62
     ACS
    -0.61
     SOC
    -0.60
     FTC
    -0.60
     NS
    -0.58
     captivity
    -0.58
     Freed
    -0.57
     Pearce
    -0.57
    POSITIVE LOGITS
    minute
    1.33
     minute
    1.11
    second
    1.06
     resort
    0.89
    Rated
    0.88
     ditch
    0.86
    chance
    0.83
     Minute
    0.82
     second
    0.77
     secondly
    0.75
    Activations Density 0.386%

    No Known Activations