© 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. 8-RES_POST_32K-OAI
    11. 5100
    Prev
    Next
    INDEX
    Explanations

    technical terms and components related to programming or computational functions

    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
    485
    +0.10
    0.5%
    496
    +0.10
    0.5%
    583
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.10
    0.06
    678
    +0.10
    0.05
    326
    +0.10
    0.07
    Negative Logits
    YL
    -0.79
     mourning
    -0.71
    externalActionCode
    -0.67
    Solar
    -0.66
     monuments
    -0.66
     Ascend
    -0.64
    Northern
    -0.63
     Anniversary
    -0.63
    BSD
    -0.63
    ashtra
    -0.61
    POSITIVE LOGITS
     inputs
    0.99
     input
    0.99
    Output
    0.98
     outputs
    0.98
    output
    0.96
     Output
    0.95
    input
    0.87
     Input
    0.85
    Input
    0.82
     output
    0.79
    Activations Density 0.686%

    No Known Activations