© 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. 8-RES_POST_32K-OAI
    11. 4338
    Prev
    Next
    INDEX
    Explanations

    personal names and identities within the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    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
    436
    +0.14
    0.6%
    218
    +0.13
    0.6%
    353
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.14
    0.13
    430
    +0.13
    0.13
    267
    +0.12
    0.08
    Negative Logits
    iners
    -0.71
    quez
    -0.71
    imize
    -0.67
    letes
    -0.67
     decreases
    -0.66
     bonuses
    -0.65
     worsen
    -0.65
     undermines
    -0.65
     minimize
    -0.65
     undercut
    -0.64
    POSITIVE LOGITS
    Born
    0.93
     Born
    0.83
    whose
    0.82
     Welcome
    0.80
    formerly
    0.80
     Founder
    0.79
    photo
    0.79
     born
    0.78
    alias
    0.76
     formerly
    0.74
    Activations Density 1.880%

    No Known Activations