© 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. 4388
    Prev
    Next
    INDEX
    Explanations

    sentences with strong emotional or impactful statements

    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
    148
    +0.12
    0.5%
    361
    +0.10
    0.5%
    330
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.12
    0.14
    145
    +0.10
    0.12
    440
    +0.10
    0.12
    Negative Logits
     negotiator
    -0.76
    reau
    -0.71
    illard
    -0.69
    VB
    -0.65
    jug
    -0.61
    cussion
    -0.61
     writes
    -0.60
    essional
    -0.59
    ptive
    -0.57
    ayer
    -0.56
    POSITIVE LOGITS
     '[
    0.79
     Rohing
    0.69
    Ku
    0.67
     pupils
    0.66
    engu
    0.65
    acha
    0.65
    atio
    0.64
    aki
    0.63
    onto
    0.63
    cha
    0.63
    Activations Density 0.750%

    No Known Activations