© 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. 10-RES_POST_32K-OAI
    11. 28235
    Prev
    Next
    INDEX
    Explanations

    references to complex narratives or plot structures

    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
    7
    +0.13
    0.6%
    339
    +0.11
    0.5%
    230
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.13
    0.16
    539
    +0.11
    0.13
    143
    +0.10
    0.14
    Negative Logits
     appre
    -0.95
     lapt
    -0.86
    jri
    -0.83
     seiz
    -0.80
    yip
    -0.80
    etooth
    -0.80
     mathemat
    -0.79
    bley
    -0.77
    gre
    -0.77
    userc
    -0.76
    POSITIVE LOGITS
     screenplay
    0.92
     Whedon
    0.91
     synopsis
    0.91
    Fans
    0.91
     Marvel
    0.90
     Rowling
    0.89
    Synopsis
    0.89
     Fans
    0.89
     Emmy
    0.88
     Ghostbusters
    0.88
    Activations Density 1.305%

    No Known Activations