© 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. 5368
    Prev
    Next
    INDEX
    Explanations

    gender-related terms and references to societal roles

    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
    630
    +0.11
    0.5%
    552
    +0.09
    0.4%
    527
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    401
    +0.11
    0.09
    288
    +0.09
    0.09
    527
    +0.09
    0.04
    Negative Logits
    reetings
    -0.69
    etheless
    -0.69
    umbn
    -0.69
    hide
    -0.66
    ĸļ
    -0.64
    apolis
    -0.64
    asma
    -0.64
    iatus
    -0.63
     ancest
    -0.62
    iably
    -0.62
    POSITIVE LOGITS
     stays
    0.73
     replaced
    0.67
     eliminated
    0.66
     fulfilled
    0.64
     responds
    0.64
     accordingly
    0.64
     unavailable
    0.63
     doesnt
    0.62
     likewise
    0.62
     won
    0.62
    Activations Density 1.282%

    No Known Activations