Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Attention Out - 128k
    5. 0-ATT_128K-OAI
    6. 6445
    Prev
    Next
    INDEX
    Explanations

    phrases related to processes and actions involving change or transformation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/GPT2-Small-OAI-v5-128k-attn-out-SAEs/v5_128k_layer_0
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    131,072
    Data Type
    torch.float32
    Hook Name
    blocks.0.hook_attn_out
    Hook Layer
    0
    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
    60
    +0.17
    1.1%
    336
    +0.17
    1.1%
    685
    +0.15
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.17
    0.11
    60
    +0.17
    0.08
    466
    +0.15
    0.07
    Negative Logits
    poke
    -0.80
    igans
    -0.75
    ppings
    -0.73
     Ancients
    -0.72
    oji
    -0.71
     railing
    -0.71
    pron
    -0.70
     gamb
    -0.69
     satir
    -0.69
    bill
    -0.69
    POSITIVE LOGITS
    hement
    0.88
     process
    0.87
    itored
    0.84
    URA
    0.83
    itionally
    0.83
     processes
    0.82
    vironment
    0.80
    MRI
    0.80
    SPONSORED
    0.79
    TION
    0.78
    Activations Density 2.659%

    No Known Activations