© 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. 6927
    Prev
    Next
    INDEX
    Explanations

    numerical values and certain keywords reflecting quantity or measures

    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
    505
    +0.10
    0.5%
    373
    +0.10
    0.5%
    431
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    650
    +0.10
    0.02
    675
    +0.10
    0.02
    168
    +0.10
    0.02
    Negative Logits
    aspers
    -0.70
    usercontent
    -0.68
    oldemort
    -0.67
     hardness
    -0.65
    arrett
    -0.65
    ÅĤ
    -0.64
    efficients
    -0.62
    uron
    -0.62
    awed
    -0.62
     swall
    -0.60
    POSITIVE LOGITS
     Contributions
    0.69
    ¯
    0.69
    ij
    0.66
     Gifts
    0.65
    ractor
    0.65
    Ĥİ
    0.65
     Gift
    0.65
     Benef
    0.62
    æĥ
    0.62
     Offensive
    0.62
    Activations Density 0.064%

    No Known Activations