© 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. 10-RES_POST_32K-OAI
    11. 28076
    Prev
    Next
    INDEX
    Explanations

    concepts related to policies and their implications

    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
    666
    +0.13
    0.6%
    31
    +0.12
    0.6%
    547
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    253
    +0.13
    0.03
    489
    +0.12
    0.03
    759
    +0.11
    0.03
    Negative Logits
    layer
    -0.70
     brill
    -0.69
    wear
    -0.68
    comings
    -0.68
    bands
    -0.67
     orders
    -0.66
    linger
    -0.65
    orders
    -0.64
     builders
    -0.63
    icka
    -0.63
    POSITIVE LOGITS
     ``
    0.73
     "_
    0.66
     thereof
    0.66
     "@
    0.60
     Trust
    0.59
     "/
    0.59
    ãĥķãĤ¡
    0.59
     `
    0.59
     '/
    0.58
     #####
    0.58
    Activations Density 0.215%

    No Known Activations