© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 10-RES_POST_32K-OAI
    6. 29501
    Prev
    Next
    INDEX
    Explanations

    the phrase "so-called" indicating skepticism or disparagement about a term or label

    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
    666
    +0.13
    0.6%
    425
    +0.12
    0.6%
    373
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    425
    +0.13
    0.02
    580
    +0.12
    0.02
    583
    +0.11
    0.02
    Negative Logits
    theless
    -0.90
     srfAttach
    -0.67
     Quote
    -0.63
     diplom
    -0.62
    ufact
    -0.61
     IPM
    -0.61
     SPACE
    -0.60
     Seasons
    -0.58
     Diary
    -0.55
     propri
    -0.53
    POSITIVE LOGITS
    -
    0.93
    called
    0.92
     called
    0.91
    âĢIJ
    0.85
    âĢij
    0.85
    vy
    0.83
    vi
    0.83
    ctic
    0.80
    irs
    0.75
    va
    0.75
    Activations Density 0.089%

    No Known Activations