© 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. 6165
    Prev
    Next
    INDEX
    Explanations

    the phrase "I do" and its variations, indicating expressions of intention or belief

    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
    +0.11
    0.5%
    536
    +0.11
    0.5%
    407
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    17
    +0.11
    0.05
    287
    +0.11
    0.05
    208
    +0.11
    0.05
    Negative Logits
     unforeseen
    -0.70
    PDATE
    -0.68
     swift
    -0.68
    oaded
    -0.65
     harsher
    -0.65
    uffle
    -0.65
     Kinnikuman
    -0.65
    ancial
    -0.63
    allery
    -0.62
     spirited
    -0.62
    POSITIVE LOGITS
     intend
    1.04
     know
    1.02
     owe
    0.99
     believe
    0.95
     understand
    0.86
     think
    0.84
     expect
    0.83
     knew
    0.82
     regretted
    0.81
     suspect
    0.79
    Activations Density 0.151%

    No Known Activations