© 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. 8-RES_POST_32K-OAI
    11. 4504
    Prev
    Next
    INDEX
    Explanations

    sources or citations in the text

    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
    373
    +0.19
    0.9%
    739
    +0.12
    0.6%
    649
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    334
    +0.19
    0.02
    373
    +0.12
    -0.01
    649
    +0.12
    0.01
    Negative Logits
    ecause
    -0.77
    inary
    -0.69
     cartridges
    -0.67
    matically
    -0.67
    antle
    -0.66
     desserts
    -0.66
    phas
    -0.65
    MpServer
    -0.64
     baskets
    -0.64
    APTER
    -0.64
    POSITIVE LOGITS
     Getty
    0.98
     Huffington
    0.89
     Wikimedia
    0.89
     Sina
    0.82
     Reuters
    0.81
     REUTERS
    0.81
     HuffPost
    0.80
     Eater
    0.77
     Fortune
    0.75
     ACLU
    0.74
    Activations Density 0.111%

    No Known Activations