© 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. 26898
    Prev
    Next
    INDEX
    Explanations

    references to immigration and deportation policies

    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
    720
    +0.13
    0.6%
    314
    +0.12
    0.5%
    1
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    720
    +0.13
    0.12
    280
    +0.12
    0.08
    86
    +0.10
    0.08
    Negative Logits
    âĵĺ
    -0.88
     Lauder
    -0.83
     Canaver
    -0.72
     embr
    -0.71
     delegation
    -0.69
    DragonMagazine
    -0.69
    english
    -0.68
     congrat
    -0.67
     delim
    -0.67
     ±
    -0.66
    POSITIVE LOGITS
    danger
    0.86
     dangerously
    0.85
     fel
    0.84
     felony
    0.84
    violent
    0.83
     pose
    0.83
     crimes
    0.82
     misdem
    0.82
     dangerous
    0.82
     endanger
    0.82
    Activations Density 3.416%

    No Known Activations