© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 4253
    Prev
    Next
    INDEX
    Explanations

    references to relative size or rank of entities, particularly with the term "second largest."

    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
    506
    +0.14
    0.7%
    460
    +0.12
    0.6%
    466
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.14
    0.04
    425
    +0.12
    0.04
    460
    +0.12
    0.03
    Negative Logits
    bos
    -0.73
    aiman
    -0.68
    EStream
    -0.65
    kees
    -0.65
    cgi
    -0.65
    AUT
    -0.64
    tnc
    -0.63
    estern
    -0.61
    thirst
    -0.59
    omic
    -0.58
    POSITIVE LOGITS
    imester
    0.95
    tier
    0.87
    generation
    0.87
    degree
    0.87
    rate
    0.76
    ranked
    0.76
    highest
    0.75
    party
    0.75
    largest
    0.75
    century
    0.73
    Activations Density 0.084%

    No Known Activations