© 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
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 10-RES_POST_32K-OAI
    11. 29303
    Prev
    Next
    INDEX
    Explanations

    intense criticism or strong reactions to political or social issues

    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
    556
    +0.14
    0.6%
    180
    +0.10
    0.5%
    606
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    616
    +0.14
    0.08
    288
    +0.10
    0.10
    374
    +0.10
    -0.03
    Negative Logits
     stupid
    -0.78
     unknow
    -0.78
     fucked
    -0.74
     laz
    -0.74
     Wonderful
    -0.73
     wonderful
    -0.73
     shitty
    -0.73
     marvelous
    -0.72
     lazy
    -0.72
     wonderfully
    -0.71
    POSITIVE LOGITS
     rebuke
    1.02
     condemnation
    0.99
     criticism
    0.99
     criticisms
    0.89
     rebutt
    0.85
     critiques
    0.85
     rhetoric
    0.83
     denounce
    0.82
     critics
    0.81
     opposition
    0.80
    Activations Density 2.021%

    No Known Activations