© 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. 10-RES_POST_32K-OAI
    11. 28322
    Prev
    Next
    INDEX
    Explanations

    sections related to community support and development initiatives

    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
    253
    +0.10
    0.5%
    50
    +0.10
    0.5%
    431
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    508
    +0.10
    0.09
    723
    +0.10
    0.07
    756
    +0.10
    0.01
    Negative Logits
     nowhere
    -0.88
     desper
    -0.87
     sinners
    -0.83
    sorry
    -0.80
    addafi
    -0.79
     dod
    -0.78
     horr
    -0.77
    dirty
    -0.76
    Sorry
    -0.76
     mysteriously
    -0.74
    POSITIVE LOGITS
     partnerships
    1.39
     outreach
    1.28
     collaborative
    1.16
     collaborations
    1.12
     initiatives
    1.10
     partnership
    1.07
     partnering
    1.06
     coordinated
    1.06
     proactive
    1.04
     collaboration
    1.04
    Activations Density 2.741%

    No Known Activations