© 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. 8780
    Prev
    Next
    INDEX
    Explanations

    terms and concepts related to decentralized finance and peer-to-peer networks

    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
    221
    +0.13
    0.6%
    373
    +0.10
    0.5%
    756
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.12
    575
    +0.10
    0.09
    660
    +0.10
    0.06
    Negative Logits
     misunder
    -0.70
    ouf
    -0.69
    acca
    -0.69
    worn
    -0.68
    verages
    -0.65
    rawdownloadcloneembedreportprint
    -0.65
    older
    -0.65
    ulet
    -0.65
    CVE
    -0.65
     Fenrir
    -0.64
    POSITIVE LOGITS
     computing
    1.03
     workplaces
    0.91
     policing
    0.88
     agriculture
    0.83
     gaming
    0.82
     warfare
    0.79
     environments
    0.78
     solutions
    0.78
     alternatives
    0.78
     Computing
    0.78
    Activations Density 1.115%

    No Known Activations