© 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. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 5529
    Prev
    Next
    INDEX
    Explanations

    references to the term "Father" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    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
    385
    +0.15
    0.7%
    114
    +0.12
    0.5%
    756
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.15
    0.02
    288
    +0.12
    0.02
    343
    +0.11
    0.01
    Negative Logits
     reel
    -0.70
    axter
    -0.66
    à¸
    -0.65
    TOP
    -0.61
     ostr
    -0.61
    TOR
    -0.59
     Albion
    -0.58
     slider
    -0.57
    eters
    -0.56
     Oo
    -0.55
    POSITIVE LOGITS
    andestine
    0.83
    gage
    0.78
    idia
    0.77
    eno
    0.73
    esian
    0.73
    stice
    0.73
    isance
    0.71
    agram
    0.71
    hedral
    0.70
    inity
    0.69
    Activations Density 0.096%

    No Known Activations