© 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. 5444
    Prev
    Next
    INDEX
    Explanations

    terms related to groups or quantities of people

    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
    218
    +0.12
    0.6%
    223
    +0.09
    0.4%
    450
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    313
    +0.12
    0.06
    159
    +0.09
    0.06
    450
    +0.09
    0.04
    Negative Logits
    advertising
    -0.90
    netflix
    -0.78
     FAA
    -0.77
    ĸļ
    -0.76
    NAS
    -0.75
    Downloadha
    -0.75
    ICO
    -0.73
    é¾įå¥ij士
    -0.73
    osterone
    -0.72
     ABV
    -0.72
    POSITIVE LOGITS
     persons
    1.08
     Greeks
    0.93
     eminent
    0.89
    morrow
    0.87
     deeds
    0.87
     philosophers
    0.87
     evils
    0.86
     pilgrims
    0.86
     sects
    0.84
     nations
    0.82
    Activations Density 0.391%

    No Known Activations