© 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 EvalsExportsAPI 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. 6594
    Prev
    Next
    INDEX
    Explanations

    phrases related to usability and functionality in various contexts

    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
    751
    +0.11
    0.5%
    418
    +0.11
    0.5%
    198
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.11
    0.15
    102
    +0.11
    0.14
    566
    +0.10
    0.10
    Negative Logits
    facts
    -0.69
    erald
    -0.65
    rawdownloadcloneembedreportprint
    -0.65
     Salmon
    -0.64
     transcript
    -0.63
     gulf
    -0.63
     redacted
    -0.61
     Boh
    -0.60
     determination
    -0.60
    avanaugh
    -0.59
    POSITIVE LOGITS
     installed
    1.07
     marketed
    1.01
     deployed
    0.99
     purchased
    0.96
     utilized
    0.95
     used
    0.93
     worn
    0.89
     discontinued
    0.89
     fielded
    0.87
     Used
    0.85
    Activations Density 1.587%

    No Known Activations