© 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. 4897
    Prev
    Next
    INDEX
    Explanations

    phrases related to documentation and regulatory compliance

    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
    150
    +0.09
    0.4%
    702
    +0.09
    0.4%
    88
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.09
    0.07
    393
    +0.09
    0.08
    187
    +0.09
    0.04
    Negative Logits
    inho
    -0.80
    stones
    -0.68
    lest
    -0.68
    liners
    -0.66
    tones
    -0.66
    tes
    -0.63
    liner
    -0.62
    factor
    -0.61
    odynam
    -0.60
    poke
    -0.60
    POSITIVE LOGITS
     receipt
    0.89
     intention
    0.70
     expenditures
    0.70
     receipts
    0.69
     accomplishments
    0.69
     each
    0.68
     anticipated
    0.67
     deductions
    0.67
     exemption
    0.66
     withdrawal
    0.66
    Activations Density 0.862%

    No Known Activations