© 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. 10-RES_POST_32K-OAI
    11. 29683
    Prev
    Next
    INDEX
    Explanations

    references to individuals and their actions or affiliations

    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
    373
    +0.10
    0.5%
    517
    +0.09
    0.4%
    650
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    139
    +0.10
    0.03
    477
    +0.09
    0.01
    502
    +0.08
    0.01
    Negative Logits
     understatement
    -0.80
    Reviewer
    -0.73
    ilial
    -0.72
    merce
    -0.70
     Worse
    -0.66
    geist
    -0.66
    ourgeois
    -0.64
    enance
    -0.62
    £ı
    -0.61
    vironment
    -0.60
    POSITIVE LOGITS
    inia
    0.72
    osponsors
    0.65
    opened
    0.62
     Variant
    0.62
     fled
    0.61
     announced
    0.59
     pledge
    0.57
     unity
    0.56
    iru
    0.56
     toured
    0.56
    Activations Density 0.072%

    No Known Activations