© 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. 4474
    Prev
    Next
    INDEX
    Explanations

    words related to manipulation or deceit

    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
    311
    +0.10
    0.5%
    178
    +0.10
    0.5%
    671
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.10
    0.03
    299
    +0.10
    0.03
    232
    +0.09
    0.03
    Negative Logits
    phrine
    -0.77
    atari
    -0.65
     redress
    -0.63
    illac
    -0.62
     cran
    -0.62
     subp
    -0.61
     exits
    -0.60
     balcon
    -0.60
     rooft
    -0.60
     vel
    -0.59
    POSITIVE LOGITS
    é¾
    0.97
    emouth
    0.73
    ãĥīãĥ©ãĤ´ãĥ³
    0.69
    vell
    0.66
    ãĤ¨ãĥ«
    0.65
    CHA
    0.64
    EMENT
    0.64
    ãĤ¦ãĤ¹
    0.63
     Shed
    0.63
    Div
    0.63
    Activations Density 0.198%

    No Known Activations