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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 4444
    Prev
    Next
    INDEX
    Explanations

    third-person pronouns and their usage in context

    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
    86
    +0.14
    0.6%
    491
    +0.10
    0.4%
    473
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    86
    +0.14
    0.07
    466
    +0.10
    0.06
    644
    +0.10
    0.05
    Negative Logits
    ?]
    -0.67
    Interested
    -0.66
    orous
    -0.62
     })
    -0.61
     *)
    -0.61
    *)
    -0.61
    Women
    -0.61
    Entry
    -0.60
    ocular
    -0.59
    >)
    -0.58
    POSITIVE LOGITS
    uyomi
    0.89
    ulhu
    0.73
    din
    0.69
    ãģ®ç
    0.68
    gat
    0.67
    bley
    0.67
    onga
    0.67
    rower
    0.64
     transitioned
    0.62
     craw
    0.61
    Activations Density 0.800%

    No Known Activations