© 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. 7198
    Prev
    Next
    INDEX
    Explanations

    verbs related to actions and instructions

    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
    416
    +0.14
    0.6%
    152
    +0.12
    0.5%
    720
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    611
    +0.14
    0.07
    604
    +0.12
    0.06
    184
    +0.10
    0.04
    Negative Logits
    ]=
    -0.57
    emale
    -0.56
    olver
    -0.56
    158
    -0.56
    mith
    -0.55
    uton
    -0.54
    157
    -0.54
     Mehran
    -0.52
    163
    -0.52
    128
    -0.52
    POSITIVE LOGITS
     yourselves
    1.37
     yourself
    1.34
     Yourself
    1.15
     your
    1.11
     Your
    0.90
    ments
    0.90
    ings
    0.89
     YOUR
    0.88
     thy
    0.86
    ingly
    0.83
    Activations Density 0.877%

    No Known Activations