© 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. 5843
    Prev
    Next
    INDEX
    Explanations

    negations or phrases indicating the absence of something

    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
    745
    +0.16
    0.7%
    309
    +0.12
    0.6%
    674
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    754
    +0.16
    0.06
    616
    +0.12
    0.06
    455
    +0.11
    0.06
    Negative Logits
    ahime
    -0.87
    existent
    -0.79
    alam
    -0.74
    alach
    -0.71
    venture
    -0.70
    ector
    -0.70
    rica
    -0.70
     century
    -0.69
    leground
    -0.68
    genre
    -0.67
    POSITIVE LOGITS
     refunds
    0.98
     warranties
    0.94
     sneaky
    0.80
     insertion
    0.79
     exceptions
    0.77
     swapping
    0.77
     surprises
    0.76
     extras
    0.74
     adjustment
    0.73
     notifications
    0.73
    Activations Density 0.266%

    No Known Activations