© 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. 27228
    Prev
    Next
    INDEX
    Explanations

    references to employment or job-related topics

    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
    314
    +0.12
    0.5%
    146
    +0.10
    0.4%
    438
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.12
    0.04
    174
    +0.10
    0.03
    620
    +0.10
    0.02
    Negative Logits
    âĢİ
    -0.62
     Kubrick
    -0.58
     redistributed
    -0.57
     bed
    -0.57
     otherwise
    -0.57
     AVG
    -0.57
     Trout
    -0.56
     Wikimedia
    -0.56
    SourceFile
    -0.55
     inmate
    -0.55
    POSITIVE LOGITS
    peror
    1.23
    achine
    1.18
    bers
    1.12
    issions
    1.10
    erald
    1.06
    obile
    1.06
    pering
    1.04
    bourg
    1.03
    blem
    1.01
    poral
    1.00
    Activations Density 0.230%

    No Known Activations