© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5679
    Prev
    Next
    INDEX
    Explanations

    punctuations and periods in the text

    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
    64
    +0.14
    0.7%
    314
    +0.12
    0.6%
    42
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.14
    0.31
    145
    +0.12
    0.23
    440
    +0.11
    0.22
    Negative Logits
     UCH
    -0.56
     Seym
    -0.56
     Sana
    -0.56
     Kod
    -0.55
     Pog
    -0.55
     Philipp
    -0.53
    aspers
    -0.53
     Pesh
    -0.53
    enger
    -0.53
    ife
    -0.53
    POSITIVE LOGITS
    SPONSORED
    0.96
    Sounds
    0.79
    thinkable
    0.74
    *:
    0.70
     bullshit
    0.70
    *.
    0.68
    yeah
    0.68
     goddamn
    0.67
    ³³
    0.65
    ³³³³
    0.64
    Activations Density 1.961%

    No Known Activations