© 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. 28185
    Prev
    Next
    INDEX
    Explanations

    references to organizations and institutions involved in monitoring or regulating social and environmental issues

    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
    139
    +0.12
    0.6%
    296
    +0.12
    0.6%
    464
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.12
    0.08
    224
    +0.12
    0.09
    506
    +0.11
    0.06
    Negative Logits
    ANGE
    -0.70
     forgetting
    -0.67
    Translation
    -0.66
    akings
    -0.65
     upside
    -0.62
    ibaba
    -0.59
    Interstitial
    -0.59
     Conclusion
    -0.58
     ECO
    -0.57
    Atl
    -0.57
    POSITIVE LOGITS
     released
    1.18
     announced
    1.15
     unveiled
    1.08
     issued
    1.07
     applauded
    1.05
     defines
    1.02
     launched
    1.00
     publishes
    0.99
     recommends
    0.99
     conducts
    0.98
    Activations Density 0.697%

    No Known Activations