© 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. 28980
    Prev
    Next
    INDEX
    Explanations

    phrases related to rights and regulations

    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
    373
    +0.15
    0.7%
    617
    +0.12
    0.6%
    7
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.15
    -0.01
    211
    +0.12
    0.01
    425
    +0.11
    0.01
    Negative Logits
    bourg
    -0.75
     Loft
    -0.69
     JFK
    -0.60
    infeld
    -0.58
    grave
    -0.58
    isc
    -0.57
     haunt
    -0.57
     Guardiola
    -0.56
     Spect
    -0.56
     Kant
    -0.56
    POSITIVE LOGITS
    fighters
    0.83
    enta
    0.74
    enic
    0.71
    eto
    0.70
    ł
    0.65
    Fighting
    0.64
    fighter
    0.61
    ãĤ¹
    0.60
    fighting
    0.60
     ACTIONS
    0.60
    Activations Density 0.087%

    No Known Activations