© 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. 10-RES_POST_32K-OAI
    6. 29736
    Prev
    Next
    INDEX
    Explanations

    words related to deception and trickery

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    350
    +0.12
    0.5%
    464
    +0.10
    0.5%
    237
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    750
    +0.12
    0.06
    123
    +0.10
    0.07
    350
    +0.10
    0.06
    Negative Logits
    urgy
    -0.68
    ophen
    -0.66
    uben
    -0.65
    ioxide
    -0.63
    utonium
    -0.62
    porary
    -0.61
    angible
    -0.59
    inction
    -0.59
    ãĤ´ãĥ³
    -0.58
    artifacts
    -0.58
    POSITIVE LOGITS
     passers
    1.23
     onlook
    1.22
     bystanders
    1.06
     audiences
    1.03
     superiors
    1.01
    ingly
    1.01
     unsuspecting
    0.99
     readers
    0.98
     investors
    0.98
     viewers
    0.97
    Activations Density 0.653%

    No Known Activations