© 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. 10-RES_POST_32K-OAI
    11. 28606
    Prev
    Next
    INDEX
    Explanations

    instances of the letter 'o' in words

    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
    667
    +0.12
    0.5%
    702
    +0.11
    0.5%
    605
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.12
    0.02
    270
    +0.11
    0.01
    340
    +0.10
    0.01
    Negative Logits
    AME
    -0.88
    ISA
    -0.87
    OC
    -0.86
    Agent
    -0.76
    OA
    -0.76
    ACP
    -0.75
    bos
    -0.74
    AC
    -0.74
    naissance
    -0.73
    IG
    -0.72
    POSITIVE LOGITS
    ãĤ§
    0.72
     ft
    0.67
    itative
    0.67
     ru
    0.67
    ptions
    0.67
    uters
    0.66
    uple
    0.66
     lime
    0.65
    ãĥ³ãĤ¸
    0.65
     fmt
    0.64
    Activations Density 0.114%

    No Known Activations