© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 5273
    Prev
    Next
    INDEX
    Explanations

    statements expressing ability or potential

    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
    373
    +0.13
    0.6%
    566
    +0.13
    0.6%
    196
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.13
    0.07
    49
    +0.13
    0.06
    718
    +0.11
    0.05
    Negative Logits
    taboola
    -0.70
    angers
    -0.69
     perish
    -0.69
    orrow
    -0.68
    wick
    -0.66
    anger
    -0.65
    breakers
    -0.64
    afort
    -0.63
     Pastebin
    -0.62
     Profit
    -0.62
    POSITIVE LOGITS
     imagine
    0.79
     assure
    0.73
     pict
    0.71
     msec
    0.69
     envision
    0.68
     listing
    0.67
     exha
    0.65
     honestly
    0.65
    EA
    0.62
    udic
    0.62
    Activations Density 0.145%

    No Known Activations