© 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. 5603
    Prev
    Next
    INDEX
    Explanations

    instances of the letter combinations 'm' and 'r' in various contexts

    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
    +0.15
    0.7%
    464
    +0.12
    0.5%
    584
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.15
    0.07
    373
    +0.12
    -0.02
    701
    +0.09
    0.04
    Negative Logits
     mosqu
    -0.75
    ategory
    -0.73
    ailability
    -0.67
     bonded
    -0.61
     crest
    -0.61
     harb
    -0.61
    Downloadha
    -0.60
    DonaldTrump
    -0.59
     phyl
    -0.59
     Nav
    -0.59
    POSITIVE LOGITS
    ysis
    0.88
    igans
    0.84
    esa
    0.79
    iku
    0.77
    aku
    0.74
    ered
    0.72
    ity
    0.71
    Aid
    0.71
    ij
    0.71
    iac
    0.69
    Activations Density 0.318%

    No Known Activations