© 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. 4918
    Prev
    Next
    INDEX
    Explanations

    references to sexual identities and orientations

    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
    299
    +0.10
    0.5%
    318
    +0.10
    0.5%
    567
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.10
    0.19
    640
    +0.10
    0.18
    107
    +0.10
    0.15
    Negative Logits
    kefeller
    -0.63
    lled
    -0.62
    ashi
    -0.61
     Krug
    -0.58
    pring
    -0.58
     optimism
    -0.58
     Olympus
    -0.58
     Ridge
    -0.57
     yesterday
    -0.57
     revamped
    -0.57
    POSITIVE LOGITS
     Examples
    1.30
     Usually
    1.09
    typically
    1.07
    Examples
    1.05
    usually
    1.03
     Often
    1.00
    often
    1.00
     Either
    1.00
     preferably
    0.99
     Generally
    0.97
    Activations Density 2.169%

    No Known Activations