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

    Neuronpedia

    Jacobian LensNEW
    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. 6125
    Prev
    Next
    INDEX
    Explanations

    numerical values, particularly time and date expressions

    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
    373
    +0.18
    0.8%
    273
    +0.10
    0.5%
    350
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    482
    +0.18
    0.05
    373
    +0.10
    -0.03
    719
    +0.10
    0.05
    Negative Logits
    panic
    -0.82
    Äĩ
    -0.70
    iple
    -0.64
    ross
    -0.63
     Poe
    -0.60
    itatively
    -0.59
    ree
    -0.59
    izon
    -0.58
    urs
    -0.58
     determination
    -0.58
    POSITIVE LOGITS
    Aug
    0.84
    Reviewed
    0.80
     hrs
    0.78
    aughs
    0.76
     Comments
    0.75
    Apr
    0.74
    Feb
    0.74
    Released
    0.72
    pedia
    0.72
    actionDate
    0.72
    Activations Density 0.375%

    No Known Activations