© 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. 10-RES_POST_32K-OAI
    11. 29390
    Prev
    Next
    INDEX
    Explanations

    terms related to research methodology and accountability

    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
    501
    +0.11
    0.5%
    512
    +0.10
    0.5%
    577
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.22
    225
    +0.10
    0.17
    762
    +0.10
    0.18
    Negative Logits
     and
    -0.95
     or
    -0.93
     AND
    -0.90
    and
    -0.86
     &
    -0.84
     et
    -0.74
    And
    -0.73
     And
    -0.69
     nor
    -0.63
    /
    -0.60
    POSITIVE LOGITS
     alike
    1.89
     respectively
    1.26
    depending
    0.81
     sectors
    0.76
     industries
    0.73
     versa
    0.72
     combo
    0.65
     interchange
    0.63
     depending
    0.60
     combination
    0.60
    Activations Density 6.040%

    No Known Activations