© 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. 31063
    Prev
    Next
    INDEX
    Explanations

    phrases related to the impact or effectiveness of actions and their consequences

    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
    690
    +0.10
    0.5%
    340
    +0.10
    0.4%
    166
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.10
    0.06
    698
    +0.10
    0.06
    690
    +0.09
    0.05
    Negative Logits
    roid
    -0.78
    ibrary
    -0.73
    sbm
    -0.71
    anian
    -0.65
    mberg
    -0.63
     Tanz
    -0.62
    milo
    -0.62
     Puppet
    -0.62
    panic
    -0.61
    ynski
    -0.61
    POSITIVE LOGITS
     harm
    1.05
     damage
    0.88
     good
    0.84
     wonders
    0.83
     beneficial
    0.80
     measurable
    0.78
     difference
    0.75
     lasting
    0.73
     justice
    0.72
     deterrence
    0.72
    Activations Density 0.748%

    No Known Activations