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. 6476
    Prev
    Next
    INDEX
    Explanations

    expressions related to boasting or bragging

    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
    167
    +0.29
    2.3%
    484
    +0.20
    1.6%
    187
    +0.19
    1.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    85
    +0.29
    0.01
    99
    +0.20
    0.01
    151
    +0.19
    0.01
    Negative Logits
    ppelin
    -0.75
     Loving
    -0.72
     Archdemon
    -0.70
     Artemis
    -0.67
    WAYS
    -0.66
     Dominion
    -0.65
     Somers
    -0.64
     Uganda
    -0.64
    á½
    -0.63
    mary
    -0.62
    POSITIVE LOGITS
     unden
    0.85
     behavi
    0.83
     citiz
    0.81
    estamp
    0.80
    amera
    0.79
    ij士
    0.76
    ulative
    0.76
    ault
    0.76
    ety
    0.75
    ebus
    0.73
    Activations Density 0.006%

    No Known Activations