© 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. 7596
    Prev
    Next
    INDEX
    Explanations

    expressions of personal development and self-reflection

    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
    241
    +0.10
    0.4%
    427
    +0.10
    0.4%
    732
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    512
    +0.10
    0.11
    393
    +0.10
    0.19
    271
    +0.09
    0.07
    Negative Logits
     scrut
    -0.68
    Alert
    -0.66
    ItemTracker
    -0.63
    ivably
    -0.62
     Chiefs
    -0.62
    eer
    -0.62
     Experts
    -0.61
     Workers
    -0.61
    asts
    -0.60
    awks
    -0.60
    POSITIVE LOGITS
     my
    1.14
     My
    0.99
     I
    0.98
    My
    0.97
     MY
    0.94
    obbies
    0.88
     me
    0.87
     myself
    0.86
    I
    0.85
    esides
    0.77
    Activations Density 5.564%

    No Known Activations