© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 7880
    Prev
    Next
    INDEX
    Explanations

    terms related to national and federal governance

    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
    496
    +0.09
    0.4%
    139
    +0.09
    0.4%
    50
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    752
    +0.09
    0.06
    667
    +0.09
    0.04
    616
    +0.09
    0.04
    Negative Logits
    maxwell
    -0.72
     Clapper
    -0.71
    paio
    -0.71
    auc
    -0.71
    ionage
    -0.68
     Scalia
    -0.68
    âĦ¢:
    -0.67
    Ħ¢
    -0.66
     Cheney
    -0.65
     SEAL
    -0.64
    POSITIVE LOGITS
    ised
    1.20
    isation
    1.18
    isations
    1.01
    ization
    0.99
    isable
    0.99
    ism
    0.87
     colours
    0.86
     councillor
    0.84
    ities
    0.84
    izes
    0.83
    Activations Density 0.176%

    No Known Activations