© 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. 28570
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "had" along with expressions of past experiences or situations

    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
    357
    +0.13
    0.6%
    26
    +0.11
    0.5%
    515
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.13
    0.04
    730
    +0.11
    0.03
    20
    +0.10
    0.03
    Negative Logits
    mson
    -0.79
    Newsletter
    -0.77
     didnt
    -0.71
    quartered
    -0.65
    JD
    -0.64
    aday
    -0.63
    æ©
    -0.63
    arning
    -0.62
    ethy
    -0.62
    MSN
    -0.61
    POSITIVE LOGITS
     dealings
    1.04
     success
    1.03
     occasion
    0.99
     trouble
    0.92
     enough
    0.90
     difficulty
    0.89
     setbacks
    0.87
     luck
    0.86
     opportunities
    0.86
     ample
    0.85
    Activations Density 0.091%

    No Known Activations