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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & 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. 6424
    Prev
    Next
    INDEX
    Explanations

    words related to the concept of being "able" or capable

    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
    554
    +0.24
    1.8%
    276
    +0.20
    1.5%
    101
    +0.17
    1.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    301
    +0.24
    0.00
    271
    +0.20
    0.01
    302
    +0.17
    0.01
    Negative Logits
    âĸ¬
    -0.78
    pport
    -0.77
    apolis
    -0.77
    udeb
    -0.76
    etime
    -0.76
     Lumpur
    -0.75
    arat
    -0.72
    ãĤ´ãĥ³
    -0.71
    neum
    -0.71
    phalt
    -0.70
    POSITIVE LOGITS
    icut
    0.83
    KEN
    0.81
    umbnails
    0.81
    âĵĺ
    0.80
    bley
    0.78
    omsday
    0.77
    BRE
    0.73
    MENT
    0.73
    EMENT
    0.73
    mental
    0.72
    Activations Density 0.005%

    No Known Activations