© 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. 5785
    Prev
    Next
    INDEX
    Explanations

    the word "then" in various contexts indicating sequential actions or conditional statements

    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
    515
    +0.13
    0.6%
    438
    +0.09
    0.4%
    486
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.13
    0.03
    718
    +0.09
    0.03
    706
    +0.09
    0.03
    Negative Logits
     Doctrine
    -0.66
     sqor
    -0.66
     Presence
    -0.63
    oller
    -0.62
     Stam
    -0.62
     anybody
    -0.59
     interruption
    -0.58
     Conan
    -0.57
     absence
    -0.57
    ieth
    -0.57
    POSITIVE LOGITS
    DragonMagazine
    0.93
     proceeded
    0.84
     followed
    0.84
    forth
    0.80
     swiftly
    0.78
    Topic
    0.73
    upon
    0.72
     promptly
    0.70
    ilateral
    0.68
    forward
    0.68
    Activations Density 0.197%

    No Known Activations