© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 8985
    Prev
    Next
    INDEX
    Explanations

    phrases related to malfunctions or failures

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    546
    +0.11
    0.5%
    40
    +0.10
    0.5%
    218
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    351
    +0.11
    0.08
    271
    +0.10
    0.07
    570
    +0.10
    0.03
    Negative Logits
    zee
    -0.67
     altru
    -0.65
     humble
    -0.64
    ourning
    -0.62
    assad
    -0.61
    izen
    -0.61
     acknowledges
    -0.61
    ithe
    -0.60
     Nine
    -0.59
    orney
    -0.59
    POSITIVE LOGITS
     contam
    0.94
     improperly
    0.93
     havoc
    0.89
     excessively
    0.88
     unnecessarily
    0.88
     overhe
    0.87
     worse
    0.87
     prematurely
    0.86
     unnecess
    0.85
     ruining
    0.85
    Activations Density 1.932%

    No Known Activations