© 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. 8928
    Prev
    Next
    INDEX
    Explanations

    mentions of "twee" or "twe" along with variations, likely related to a specific cultural or stylistic context

    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
    420
    +0.11
    0.5%
    23
    +0.09
    0.4%
    555
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    555
    +0.11
    0.02
    23
    +0.09
    0.02
    325
    +0.09
    0.02
    Negative Logits
    inates
    -0.78
    inating
    -0.76
    ãĥĩ
    -0.76
    inated
    -0.75
    inate
    -0.65
    ITY
    -0.63
     Passage
    -0.61
    senal
    -0.60
     Crypt
    -0.60
    ination
    -0.59
    POSITIVE LOGITS
    eden
    1.06
    ety
    0.97
    ritten
    0.96
    akable
    0.95
    ollen
    0.94
    sburg
    0.92
    riter
    0.91
    ez
    0.86
    oria
    0.86
    aken
    0.85
    Activations Density 0.136%

    No Known Activations