© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 10-RES_POST_32K-OAI
    6. 29454
    Prev
    Next
    INDEX
    Explanations

    the word "there" and its various forms in context

    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
    +0.13
    0.6%
    373
    +0.12
    0.5%
    672
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    287
    +0.13
    0.17
    757
    +0.12
    0.10
    213
    +0.11
    0.10
    Negative Logits
    ãĥĵ
    -0.82
     Skydragon
    -0.79
    ãĤ¢ãĥ«
    -0.78
    Delivery
    -0.74
     cheapest
    -0.70
    ãĥĺ
    -0.69
    Zip
    -0.68
    FTWARE
    -0.67
     10000
    -0.66
     2400
    -0.66
    POSITIVE LOGITS
     certainly
    1.12
     clearly
    1.01
     appears
    0.96
     testified
    0.94
     plainly
    0.93
     surely
    0.91
     undoubtedly
    0.90
     evidence
    0.89
     seems
    0.89
     circumst
    0.88
    Activations Density 0.275%

    No Known Activations