© 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. 27963
    Prev
    Next
    INDEX
    Explanations

    mentions of projects and their statuses or characteristics

    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
    459
    +0.17
    0.8%
    651
    +0.13
    0.6%
    247
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    459
    +0.17
    0.03
    651
    +0.13
    0.04
    247
    +0.11
    0.03
    Negative Logits
    osate
    -0.89
    gel
    -0.69
    tesque
    -0.68
    dos
    -0.65
    ocaust
    -0.65
     Ange
    -0.64
    mson
    -0.64
    weed
    -0.63
    fed
    -0.62
    NetMessage
    -0.62
    POSITIVE LOGITS
     Gutenberg
    1.16
    ors
    1.11
     undertaken
    0.94
    ions
    0.90
     underway
    0.85
     manager
    0.85
    ive
    0.82
    ivities
    0.81
    ivity
    0.81
    mable
    0.79
    Activations Density 0.140%

    No Known Activations