© 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. 8-RES_POST_32K-OAI
    11. 8993
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate actions to modify, reflect, or update information

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    218
    +0.11
    0.5%
    312
    +0.11
    0.5%
    655
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.11
    0.07
    508
    +0.11
    0.04
    665
    +0.10
    0.04
    Negative Logits
    awar
    -0.66
    isner
    -0.65
    ibrary
    -0.61
    wordpress
    -0.60
    ebook
    -0.59
     dearly
    -0.59
    ictive
    -0.58
     learnt
    -0.58
    downs
    -0.57
    rontal
    -0.57
    POSITIVE LOGITS
     reflect
    1.07
     accommodate
    0.98
     Include
    0.84
     include
    0.81
     suit
    0.81
    reflect
    0.81
     conform
    0.81
     fit
    0.78
     resemble
    0.76
     incorporate
    0.68
    Activations Density 0.401%

    No Known Activations