© 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. 10-RES_POST_32K-OAI
    11. 30063
    Prev
    Next
    INDEX
    Explanations

    phrases related to terms of service and user agreements

    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
    516
    +0.12
    0.5%
    373
    +0.12
    0.5%
    197
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.12
    0.05
    685
    +0.12
    0.03
    551
    +0.11
    0.04
    Negative Logits
    sbm
    -0.71
     dstg
    -0.69
    Roaming
    -0.68
    ojure
    -0.65
     srfAttach
    -0.65
     Cutting
    -0.64
    bilt
    -0.63
    itsch
    -0.60
    itcher
    -0.60
     lun
    -0.60
    POSITIVE LOGITS
     endeavour
    0.88
     reserves
    0.83
     disclaim
    0.77
     reserve
    0.77
    ĪĴ
    0.74
     endeavor
    0.72
     assume
    0.72
     guarant
    0.71
     strive
    0.67
    orage
    0.67
    Activations Density 0.499%

    No Known Activations