© 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. 4588
    Prev
    Next
    INDEX
    Explanations

    phrases relating to timeframes and deadlines

    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
    403
    +0.11
    0.5%
    47
    +0.11
    0.5%
    512
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.11
    0.11
    16
    +0.11
    0.06
    123
    +0.11
    0.10
    Negative Logits
    grave
    -0.58
    ensitive
    -0.57
     Ancient
    -0.55
    allic
    -0.54
    emy
    -0.54
     Cipher
    -0.54
    cial
    -0.53
    alph
    -0.53
    MD
    -0.53
    itiz
    -0.53
    POSITIVE LOGITS
     seconds
    1.29
     Seconds
    1.20
     minutes
    1.18
    seconds
    1.13
     blink
    1.13
     hours
    1.09
     hour
    1.06
     milliseconds
    1.04
    hours
    1.04
     days
    1.01
    Activations Density 1.714%

    No Known Activations