© 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 EvalsExportsAPI 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. 30670
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing existence and presence within narratives

    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
    512
    +0.14
    0.7%
    91
    +0.12
    0.6%
    689
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    20
    +0.14
    0.09
    91
    +0.12
    0.08
    512
    +0.11
    0.08
    Negative Logits
    atform
    -0.79
     adjourn
    -0.76
    communication
    -0.75
     remaining
    -0.75
    delay
    -0.72
    emate
    -0.71
    cknow
    -0.70
    icipated
    -0.67
     interven
    -0.65
    ether
    -0.65
    POSITIVE LOGITS
     definitely
    1.12
     certainly
    1.09
     loads
    0.92
     surely
    0.90
     literally
    0.89
     absolutely
    0.85
     something
    0.82
     undoubtedly
    0.81
     probably
    0.79
     tons
    0.78
    Activations Density 0.293%

    No Known Activations