© 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. 10-RES_POST_32K-OAI
    11. 27682
    Prev
    Next
    INDEX
    Explanations

    instances of the word "no" and variations of "yes" or "yes."

    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
    137
    +0.10
    0.5%
    93
    +0.09
    0.4%
    164
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.10
    0.03
    659
    +0.09
    0.03
    232
    +0.08
    0.04
    Negative Logits
    mington
    -0.83
    yg
    -0.70
    ierrez
    -0.69
    ml
    -0.69
    rg
    -0.68
    ERG
    -0.67
    mble
    -0.67
    ibrary
    -0.66
    byn
    -0.65
    eness
    -0.64
    POSITIVE LOGITS
    issance
    0.72
    DOS
    0.65
    ÅĤ
    0.64
    Timeout
    0.63
     DeV
    0.61
    Corp
    0.61
     Schwar
    0.60
    Med
    0.60
    ön
    0.59
    Cola
    0.59
    Activations Density 0.254%

    No Known Activations