© 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 EvalsExportsinterp-engineNEWAPI 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. 8384
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "number" followed by a digit

    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
    +0.11
    0.5%
    236
    +0.10
    0.5%
    210
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    596
    +0.11
    0.04
    647
    +0.10
    0.03
    609
    +0.09
    0.03
    Negative Logits
    selves
    -0.63
     guid
    -0.61
     otherwise
    -0.61
    OOL
    -0.60
     Else
    -0.59
     Detected
    -0.59
     Laun
    -0.58
    orporated
    -0.57
     Modified
    -0.57
    arta
    -0.56
    POSITIVE LOGITS
    less
    0.89
    winner
    0.80
     crunch
    0.80
    lines
    0.76
    eight
    0.75
    Eight
    0.75
    playing
    0.71
    pants
    0.70
    lessness
    0.69
    four
    0.69
    Activations Density 0.272%

    No Known Activations