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. 13551
    Prev
    Next
    INDEX
    Explanations

    references to political deals or agreements

    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
    339
    +0.32
    2.5%
    549
    +0.23
    1.8%
    271
    +0.16
    1.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    549
    +0.32
    0.01
    241
    +0.23
    0.01
    247
    +0.16
    0.01
    Negative Logits
    ICAN
    -0.96
    ixed
    -0.79
    OIL
    -0.76
    Leaks
    -0.75
    aps
    -0.74
    minus
    -0.73
    amous
    -0.73
    âĹ¼
    -0.72
    etus
    -0.72
    ortion
    -0.71
    POSITIVE LOGITS
    racuse
    1.06
    sembly
    1.00
    paio
    0.98
    grounds
    0.98
     Leilan
    0.95
    theless
    0.91
    irlf
    0.88
     Chaser
    0.87
    etheless
    0.84
    dayName
    0.78
    Activations Density 0.008%

    No Known Activations