© 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. 8-RES_POST_32K-OAI
    11. 6073
    Prev
    Next
    INDEX
    Explanations

    references to specific individuals or organizations, particularly in a political context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    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
    191
    +0.11
    0.5%
    42
    +0.11
    0.5%
    314
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.11
    0.04
    760
    +0.11
    0.02
    409
    +0.09
    0.02
    Negative Logits
     Entered
    -0.83
     Augusta
    -0.73
     Bend
    -0.71
     Wat
    -0.70
     Shel
    -0.70
     Dimension
    -0.68
     Sim
    -0.67
     Maxim
    -0.64
     Circ
    -0.63
     Gam
    -0.61
    POSITIVE LOGITS
    psc
    0.81
    tumblr
    0.78
    govtrack
    0.77
    album
    0.71
    Tea
    0.71
    administ
    0.70
    WER
    0.69
    cffffcc
    0.68
    olkien
    0.67
    multi
    0.66
    Activations Density 0.155%

    No Known Activations