© 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. 5353
    Prev
    Next
    INDEX
    Explanations

    possessive forms of nouns

    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
    195
    +0.11
    0.5%
    65
    +0.11
    0.5%
    598
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    15
    +0.11
    0.06
    499
    +0.11
    0.05
    66
    +0.09
    0.06
    Negative Logits
    Ú
    -0.82
    Ò
    -0.80
    =]
    -0.80
    âľ
    -0.75
    Ïī
    -0.74
    Ô
    -0.74
    Õ
    -0.72
    EEE
    -0.72
    /,
    -0.71
    ̶
    -0.70
    POSITIVE LOGITS
     spokesman
    0.93
     proponents
    0.92
     detractors
    0.92
     aim
    0.90
     lawyers
    0.88
     editors
    0.88
     critics
    0.86
     spokesperson
    0.86
     spokeswoman
    0.86
     motive
    0.86
    Activations Density 0.241%

    No Known Activations