© 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. 28094
    Prev
    Next
    INDEX
    Explanations

    words related to exclusion and criteria for being left out

    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
    710
    +0.12
    0.5%
    289
    +0.10
    0.5%
    681
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    691
    +0.12
    0.02
    204
    +0.10
    0.02
    225
    +0.10
    0.02
    Negative Logits
    ãĥ¥
    -0.76
    oÄŁ
    -0.74
     upgr
    -0.72
    leeve
    -0.71
    hare
    -0.65
     Simulator
    -0.64
    ©
    -0.64
    nai
    -0.63
     ..............
    -0.63
    lycer
    -0.62
    POSITIVE LOGITS
    minded
    0.90
     criteria
    0.81
    ively
    0.76
    entials
    0.76
     altogether
    0.75
     minded
    0.75
    agne
    0.72
    ives
    0.71
    ary
    0.69
    ivity
    0.69
    Activations Density 0.194%

    No Known Activations