© 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
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 5588
    Prev
    Next
    INDEX
    Explanations

    commas and punctuation marks indicating separations or pauses in 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
    55
    +0.11
    0.5%
    132
    +0.10
    0.5%
    471
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    43
    +0.11
    0.10
    546
    +0.10
    0.11
    720
    +0.10
    0.12
    Negative Logits
     Cosponsors
    -0.76
    License
    -0.69
    iculty
    -0.69
     parap
    -0.68
    onyms
    -0.67
     resemblance
    -0.67
     recy
    -0.67
    NRS
    -0.64
     resemb
    -0.62
    ran
    -0.62
    POSITIVE LOGITS
     October
    1.13
     January
    1.12
     December
    1.11
     2019
    1.11
    months
    1.10
     timeframe
    1.09
     months
    1.08
     September
    1.08
     February
    1.06
     November
    1.06
    Activations Density 1.009%

    No Known Activations