© 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. 8-RES_POST_32K-OAI
    11. 6160
    Prev
    Next
    INDEX
    Explanations

    references to local festivals and community events

    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
    415
    +0.10
    0.5%
    596
    +0.10
    0.4%
    72
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    526
    +0.10
    -0.11
    408
    +0.10
    0.22
    389
    +0.10
    0.07
    Negative Logits
     embryos
    -0.93
     Clapper
    -0.88
     retracted
    -0.86
     actu
    -0.85
    illus
    -0.81
    heed
    -0.80
     implant
    -0.79
     FISA
    -0.79
     embry
    -0.79
     inert
    -0.78
    POSITIVE LOGITS
     locals
    1.60
     downtown
    1.43
     tourists
    1.41
     tourist
    1.37
     bustling
    1.32
    owntown
    1.31
     Downtown
    1.29
     regulars
    1.28
     town
    1.26
     cafes
    1.23
    Activations Density 12.124%

    No Known Activations