© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 10-RES_POST_32K-OAI
    11. 26756
    Prev
    Next
    INDEX
    Explanations

    phrases that describe the state or condition of an object or situation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    218
    +0.14
    0.6%
    512
    +0.11
    0.5%
    95
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    512
    +0.14
    0.11
    55
    +0.11
    0.12
    7
    +0.10
    0.09
    Negative Logits
    ranch
    -0.90
    clus
    -0.75
    mson
    -0.74
    busters
    -0.73
    quote
    -0.72
    riber
    -0.70
    soever
    -0.68
     Tycoon
    -0.65
    redit
    -0.65
    cript
    -0.64
    POSITIVE LOGITS
     unbearable
    1.16
     excruciating
    1.00
     faint
    0.99
     blinding
    0.98
     indistinguishable
    0.96
     intolerable
    0.92
     unsettling
    0.91
     horrifying
    0.89
     frightening
    0.89
     eerie
    0.89
    Activations Density 1.179%

    No Known Activations