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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 13524
    Prev
    Next
    INDEX
    Explanations

    references to task forces involved in administrative or research activities

    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
    16
    +0.27
    1.8%
    255
    +0.18
    1.1%
    503
    +0.17
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    200
    +0.27
    0.07
    507
    +0.18
    0.06
    255
    +0.17
    0.01
    Negative Logits
    nings
    -0.82
    aunder
    -0.74
     trademarks
    -0.72
    gio
    -0.72
    ãĤ¦ãĤ¹
    -0.71
    cott
    -0.70
    76561
    -0.69
    bol
    -0.68
     BALL
    -0.68
    cloth
    -0.67
    POSITIVE LOGITS
     task
    1.03
    achus
    0.95
    abulary
    0.90
    ezvous
    0.81
    RESULTS
    0.78
     cumbers
    0.77
    anooga
    0.76
    idates
    0.76
     exha
    0.74
    otor
    0.73
    Activations Density 3.147%

    No Known Activations