© 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 EvalsExportsinterp-engineNEWAPI 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. 31583
    Prev
    Next
    INDEX
    Explanations

    programming constructs and templates related to data structures, particularly in a language like C++

    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
    373
    +0.13
    0.6%
    145
    +0.11
    0.5%
    83
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.13
    0.05
    176
    +0.11
    0.02
    399
    +0.11
    0.03
    Negative Logits
    conservancy
    -0.91
    xus
    -0.90
    Downloadha
    -0.89
    spir
    -0.84
    urga
    -0.84
    vous
    -0.82
    ãĤ¦ãĤ¹
    -0.81
    agall
    -0.80
    ãĥ¯ãĥ³
    -0.80
    tun
    -0.78
    POSITIVE LOGITS
     Main
    0.89
     bool
    0.86
     Output
    0.82
     Example
    0.82
     File
    0.80
     Begin
    0.79
     Test
    0.78
     Keyboard
    0.78
     void
    0.77
     Debug
    0.77
    Activations Density 0.251%

    No Known Activations