© 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. 4858
    Prev
    Next
    INDEX
    Explanations

    references to rituals or ceremonial 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
    288
    +0.13
    0.6%
    373
    +0.12
    0.6%
    111
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.02
    526
    +0.12
    0.00
    111
    +0.12
    0.01
    Negative Logits
    restling
    -0.73
    eco
    -0.70
    ability
    -0.66
    ordan
    -0.65
    yk
    -0.65
    eenth
    -0.64
    itar
    -0.64
    oglu
    -0.63
    sth
    -0.62
    itarian
    -0.62
    POSITIVE LOGITS
    ificial
    0.97
    urations
    0.79
     altar
    0.77
    bris
    0.76
    ifice
    0.74
     maiden
    0.73
    urn
    0.72
    ãħĭ
    0.66
     ashes
    0.65
    osite
    0.65
    Activations Density 0.044%

    No Known Activations