© 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. 6395
    Prev
    Next
    INDEX
    Explanations

    terms related to physical or conceptual boundaries

    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%
    667
    +0.10
    0.4%
    183
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.04
    325
    +0.10
    0.03
    537
    +0.09
    0.03
    Negative Logits
    milo
    -0.81
    orah
    -0.80
    ocker
    -0.76
    imar
    -0.75
    ulz
    -0.73
    atility
    -0.72
    hur
    -0.72
    akia
    -0.70
    ived
    -0.69
    otal
    -0.68
    POSITIVE LOGITS
     boundaries
    0.90
     separating
    0.86
     boundary
    0.84
     delim
    0.81
     markers
    0.79
     deline
    0.77
     clearance
    0.77
     walls
    0.76
     dividing
    0.75
     Governments
    0.74
    Activations Density 0.109%

    No Known Activations