© 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. 5937
    Prev
    Next
    INDEX
    Explanations

    expressions related to actions and time-related sequences

    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
    662
    +0.13
    0.6%
    150
    +0.10
    0.4%
    154
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    300
    +0.13
    0.04
    274
    +0.10
    0.04
    460
    +0.09
    0.04
    Negative Logits
    pill
    -0.71
    absolutely
    -0.70
     Vide
    -0.68
    ocry
    -0.65
    gow
    -0.61
    repeat
    -0.60
     Called
    -0.59
    ervatives
    -0.58
    imental
    -0.58
    wives
    -0.57
    POSITIVE LOGITS
     ensures
    0.67
     entails
    0.60
     creates
    0.59
     brings
    0.58
     istg
    0.56
    asis
    0.56
     sear
    0.55
    ר
    0.55
    iss
    0.54
    lass
    0.54
    Activations Density 0.409%

    No Known Activations