© 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. 4801
    Prev
    Next
    INDEX
    Explanations

    expressions of skepticism or criticism towards individuals or practices

    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
    745
    +0.13
    0.6%
    41
    +0.12
    0.5%
    51
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    41
    +0.13
    0.05
    49
    +0.12
    0.04
    737
    +0.11
    0.04
    Negative Logits
    utenant
    -0.70
    neutral
    -0.68
    forth
    -0.64
    erity
    -0.64
    raltar
    -0.63
     vacated
    -0.63
    agher
    -0.61
    UTH
    -0.59
    planes
    -0.57
    bil
    -0.57
    POSITIVE LOGITS
     compared
    0.73
    Feel
    0.72
    mouth
    0.67
    inventoryQuantity
    0.67
     sounding
    0.64
    DOC
    0.59
    arger
    0.58
    Laughs
    0.58
     mentally
    0.58
     feeling
    0.56
    Activations Density 0.622%

    No Known Activations