© 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. 7437
    Prev
    Next
    INDEX
    Explanations

    specific dates indicated in the text

    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
    373
    +0.17
    0.8%
    125
    +0.11
    0.5%
    464
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    430
    +0.17
    0.08
    127
    +0.11
    0.05
    325
    +0.10
    0.06
    Negative Logits
     ballistic
    -0.81
     penetrated
    -0.79
     interfering
    -0.74
     interfered
    -0.74
     interrupt
    -0.73
     disinteg
    -0.72
     interrupted
    -0.72
     intercepted
    -0.72
     notwithstanding
    -0.71
    etheless
    -0.71
    POSITIVE LOGITS
    2018
    0.91
     2018
    0.87
     Benefit
    0.85
     Leaves
    0.81
     Winners
    0.80
     Website
    0.78
     Authors
    0.76
     Views
    0.76
    Buy
    0.76
    imon
    0.75
    Activations Density 0.291%

    No Known Activations