© 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 EvalsExportsAPI 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. 6763
    Prev
    Next
    INDEX
    Explanations

    references to government officials and senators

    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
    314
    +0.14
    0.7%
    238
    +0.12
    0.5%
    456
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.14
    0.03
    456
    +0.12
    0.02
    538
    +0.11
    0.02
    Negative Logits
     maintenance
    -0.82
     upkeep
    -0.81
     policing
    -0.72
     reinforcement
    -0.70
     management
    -0.67
     Mysteries
    -0.66
     phys
    -0.65
     Britann
    -0.65
     superst
    -0.65
     roller
    -0.64
    POSITIVE LOGITS
    ial
    1.05
     Grassley
    0.94
    ially
    0.92
    ials
    0.90
    arians
    0.86
    ileaks
    0.85
     Sanders
    0.80
    acebook
    0.79
     Schumer
    0.79
    osponsors
    0.78
    Activations Density 0.062%

    No Known Activations