© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5669
    Prev
    Next
    INDEX
    Explanations

    phrases comparing or contrasting different locations or examples

    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
    122
    +0.15
    0.7%
    650
    +0.09
    0.4%
    74
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.15
    0.12
    614
    +0.09
    0.05
    122
    +0.09
    0.07
    Negative Logits
    Description
    -0.73
     allele
    -0.64
    icer
    -0.64
     antibodies
    -0.63
    progress
    -0.62
     efficacy
    -0.62
    VIDIA
    -0.60
    Decre
    -0.60
    ebus
    -0.60
     trem
    -0.59
    POSITIVE LOGITS
     Mecca
    0.95
     academia
    0.94
     Cannes
    0.90
     caves
    0.90
     Wonderland
    0.89
     airports
    0.89
     coasts
    0.88
     nowhere
    0.87
     Borders
    0.87
     heaven
    0.86
    Activations Density 2.002%

    No Known Activations