© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 10-RES_POST_32K-OAI
    11. 31369
    Prev
    Next
    INDEX
    Explanations

    phrases related to varying conditions or contexts that influence outcomes

    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
    508
    +0.12
    0.5%
    41
    +0.10
    0.4%
    214
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    41
    +0.12
    0.05
    516
    +0.10
    0.04
    508
    +0.10
    0.04
    Negative Logits
     useless
    -0.76
     worthless
    -0.72
     meaningless
    -0.71
     forever
    -0.71
     indefinitely
    -0.71
     pointless
    -0.70
     Advertisement
    -0.70
    unless
    -0.69
     harmless
    -0.69
    ????????
    -0.68
    POSITIVE LOGITS
     circumstance
    0.80
     circumstances
    0.79
     chosen
    0.78
     situation
    0.69
     locale
    0.69
     selected
    0.67
    uncture
    0.65
    settings
    0.64
     availability
    0.64
    ractor
    0.64
    Activations Density 1.281%

    No Known Activations