© 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. 27248
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing the concept of totality or completeness

    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
    20
    +0.11
    0.5%
    44
    +0.10
    0.5%
    237
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    88
    +0.11
    0.02
    484
    +0.10
    0.02
    51
    +0.09
    0.02
    Negative Logits
    stasy
    -0.71
    keeping
    -0.70
    utral
    -0.70
     deletion
    -0.68
    itiz
    -0.66
    gery
    -0.65
    ahu
    -0.64
    venge
    -0.63
    76561
    -0.62
    cko
    -0.60
    POSITIVE LOGITS
     costs
    1.19
     Costs
    1.13
     times
    1.03
    cost
    0.96
     hazards
    0.91
     cost
    0.90
     stages
    0.85
     Rates
    0.82
    Args
    0.82
    æľ
    0.81
    Activations Density 0.325%

    No Known Activations