© 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. 4320
    Prev
    Next
    INDEX
    Explanations

    content related to infectious diseases and their transmission

    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
    265
    +0.12
    0.5%
    483
    +0.09
    0.4%
    655
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    742
    +0.12
    0.08
    127
    +0.09
    0.06
    252
    +0.09
    0.09
    Negative Logits
    ĸļ
    -0.91
    iture
    -0.89
    cknow
    -0.84
    selage
    -0.84
    natureconservancy
    -0.80
    cean
    -0.77
    )=(
    -0.76
     pedest
    -0.76
    Reviewer
    -0.76
    fram
    -0.74
    POSITIVE LOGITS
     virus
    1.69
     infection
    1.68
     infections
    1.63
     pathogens
    1.57
     viruses
    1.54
     Virus
    1.48
     bacteria
    1.45
    infect
    1.45
     toxin
    1.45
     bacter
    1.41
    Activations Density 2.179%

    No Known Activations