© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5840
    Prev
    Next
    INDEX
    Explanations

    terminology related to diet and healthy living

    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
    575
    +0.13
    0.6%
    292
    +0.10
    0.4%
    225
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.13
    0.10
    225
    +0.10
    0.07
    288
    +0.09
    0.09
    Negative Logits
     wiret
    -0.76
    orescence
    -0.71
    Ni
    -0.70
    orescent
    -0.70
    ensen
    -0.67
     Qiao
    -0.67
     Citation
    -0.66
    ulz
    -0.65
    adium
    -0.65
    roy
    -0.65
    POSITIVE LOGITS
     habits
    1.25
     regimen
    1.25
     lifestyle
    1.18
     routines
    0.92
    ifestyle
    0.92
     behaviors
    0.89
     behaviours
    0.88
     choices
    0.88
     routine
    0.85
     Eating
    0.82
    Activations Density 0.715%

    No Known Activations