© 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. 4188
    Prev
    Next
    INDEX
    Explanations

    instances of statistics or numerical data related to performance and evaluations

    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
    +0.11
    0.5%
    295
    +0.10
    0.5%
    218
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    314
    +0.11
    0.03
    346
    +0.10
    0.05
    412
    +0.10
    0.03
    Negative Logits
    )].
    -0.74
    ractions
    -0.70
    Secondly
    -0.68
    wreck
    -0.66
    ryce
    -0.65
    iasco
    -0.61
     Ley
    -0.61
     srfAttach
    -0.61
     McCoy
    -0.60
    ittee
    -0.60
    POSITIVE LOGITS
    anon
    0.66
     LinkedIn
    0.66
    ä
    0.65
    ginx
    0.65
    upload
    0.63
     Pinterest
    0.63
    ########
    0.62
    hash
    0.62
    asta
    0.61
    XY
    0.60
    Activations Density 0.928%

    No Known Activations