© 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. 5970
    Prev
    Next
    INDEX
    Explanations

    sequences that indicate duration or temporal relationships

    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
    31
    +0.15
    0.7%
    494
    +0.12
    0.6%
    477
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    718
    +0.15
    0.04
    608
    +0.12
    0.05
    139
    +0.10
    0.04
    Negative Logits
    soType
    -0.75
    edIn
    -0.69
    tur
    -0.68
     secrecy
    -0.68
    ħĭ
    -0.66
    çīĪ
    -0.64
    IDA
    -0.63
     avoidance
    -0.63
    ãĥı
    -0.63
    taking
    -0.61
    POSITIVE LOGITS
     realize
    0.91
     realise
    0.82
     fully
    0.80
     reach
    0.79
     finalized
    0.79
     turnaround
    0.75
     realized
    0.75
     perfected
    0.74
     reached
    0.74
     recover
    0.71
    Activations Density 0.180%

    No Known Activations