© 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. Attention Out - 32k
    9. 3-ATT_32K-OAI
    11. 3359
    Prev
    Next
    INDEX
    Explanations

    instances of the word "wait" and its variations in conjunction with time or pauses

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-attn-out-SAEs/v5_32k_layer_3
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.3.hook_attn_out
    Hook Layer
    3
    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
    497
    +0.16
    0.8%
    3
    +0.12
    0.5%
    353
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    474
    +0.16
    0.03
    75
    +0.12
    0.04
    573
    +0.11
    0.03
    Negative Logits
    ishable
    -0.90
    20439
    -0.90
    reet
    -0.81
    è¦ļéĨĴ
    -0.80
    enta
    -0.79
    furt
    -0.78
    ented
    -0.75
    ioxide
    -0.73
    enable
    -0.72
    raught
    -0.72
    POSITIVE LOGITS
    !]
    0.82
    ?]
    0.72
     slowdown
    0.71
     enlarg
    0.67
     Didn
    0.66
    mite
    0.66
    âĢ¦]
    0.66
     typo
    0.65
     steroids
    0.64
    ]-
    0.63
    Activations Density 0.672%

    No Known Activations