© 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. 8-RES_POST_32K-OAI
    11. 5050
    Prev
    Next
    INDEX
    Explanations

    prominently featured people or entities in media and news contexts

    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
    373
    +0.12
    0.5%
    679
    +0.10
    0.4%
    469
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    386
    +0.12
    0.01
    711
    +0.10
    0.00
    120
    +0.10
    0.01
    Negative Logits
    icum
    -0.69
    ston
    -0.68
    terness
    -0.68
     helicop
    -0.67
    Closure
    -0.65
    cens
    -0.65
     sparing
    -0.65
     accur
    -0.64
    ijk
    -0.64
    eat
    -0.62
    POSITIVE LOGITS
    ãĤ¸
    0.70
     Preferred
    0.65
     Occupations
    0.64
     Videos
    0.64
     Melee
    0.63
    Prev
    0.63
     Melissa
    0.63
    Recommend
    0.61
     Previous
    0.61
    ophe
    0.59
    Activations Density 0.035%

    No Known Activations