© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 3680
    Prev
    Next
    INDEX
    Explanations

    phrases related to conclusions and summarizing statements

    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
    65
    +0.10
    0.4%
    239
    +0.09
    0.4%
    94
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    195
    +0.10
    0.03
    498
    +0.09
    0.02
    171
    +0.09
    0.02
    Negative Logits
     alike
    -0.72
     UD
    -0.69
     ali
    -0.69
     existed
    -0.65
     doub
    -0.64
     Reincarn
    -0.64
     perceived
    -0.63
     lod
    -0.63
     Franch
    -0.61
     nicknamed
    -0.61
    POSITIVE LOGITS
    ileaks
    0.85
    alore
    0.78
    autions
    0.77
    Conclusion
    0.76
    ridor
    0.75
    statement
    0.75
     concluding
    0.74
     conclusion
    0.74
    anza
    0.73
    adiq
    0.71
    Activations Density 0.601%

    No Known Activations