© 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. 10-RES_POST_32K-OAI
    11. 28927
    Prev
    Next
    INDEX
    Explanations

    references to formal documents and reports

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    +0.18
    0.8%
    197
    +0.10
    0.5%
    484
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    484
    +0.18
    0.06
    341
    +0.10
    0.05
    419
    +0.10
    0.06
    Negative Logits
    afety
    -0.66
    ystem
    -0.64
    peak
    -0.63
     mortals
    -0.59
    ãĤ¤ãĥĪ
    -0.59
    âĹ¼
    -0.59
    icum
    -0.58
    "},"
    -0.58
    successfully
    -0.57
     Birthday
    -0.57
    POSITIVE LOGITS
     reads
    0.89
     continues
    0.86
     cites
    0.85
     authors
    0.82
     warns
    0.82
     adds
    0.82
     states
    0.80
     quoted
    0.80
     cited
    0.79
     concludes
    0.78
    Activations Density 0.639%

    No Known Activations