© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    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. 6340
    Prev
    Next
    INDEX
    Explanations

    phrases related to waiting and anticipation

    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
    695
    +0.09
    0.4%
    275
    +0.09
    0.4%
    683
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    202
    +0.09
    0.02
    526
    +0.09
    -0.01
    575
    +0.09
    0.04
    Negative Logits
    çīĪ
    -0.65
    ãĥĦ
    -0.64
    ansk
    -0.64
    tein
    -0.63
     retaining
    -0.61
     fabric
    -0.60
     retained
    -0.59
     Kend
    -0.58
     Span
    -0.57
     cultivating
    -0.56
    POSITIVE LOGITS
     arrival
    1.03
     confirmation
    0.85
     reinforcements
    0.85
     arrives
    0.84
     arrive
    0.81
     opportunity
    0.80
     onset
    0.78
     replies
    0.75
    irmation
    0.75
     reply
    0.74
    Activations Density 0.574%

    No Known Activations