© 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. 30892
    Prev
    Next
    INDEX
    Explanations

    references to corporate entities and their business activities

    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
    679
    +0.10
    0.5%
    31
    +0.10
    0.5%
    436
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    139
    +0.10
    0.14
    551
    +0.10
    0.11
    86
    +0.10
    0.09
    Negative Logits
    hemer
    -0.78
    orians
    -0.71
    soDeliveryDate
    -0.71
    egal
    -0.69
     "$:/
    -0.69
    izophren
    -0.66
    VID
    -0.66
    gif
    -0.66
     forgetting
    -0.65
     remem
    -0.64
    POSITIVE LOGITS
     subsidiaries
    1.35
     IPO
    1.29
     manufacture
    1.25
     shareholders
    1.22
     manufactures
    1.18
     subcontract
    1.18
     subsidiary
    1.12
     shareholder
    1.12
     profits
    1.11
     marketed
    1.10
    Activations Density 1.545%

    No Known Activations