© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 5976
    Prev
    Next
    INDEX
    Explanations

    references to specific legal cases and their outcomes

    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
    527
    +0.10
    0.5%
    209
    +0.09
    0.4%
    680
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    756
    +0.10
    -0.04
    551
    +0.09
    0.04
    55
    +0.09
    0.07
    Negative Logits
     sacrific
    -0.80
    unts
    -0.73
     pow
    -0.73
     agre
    -0.72
    chieve
    -0.72
    istance
    -0.71
     contrace
    -0.70
     wearer
    -0.68
     gratification
    -0.67
    urances
    -0.64
    POSITIVE LOGITS
     illustrates
    0.88
     moot
    0.87
     illustrate
    0.86
     unsolved
    0.81
     Prosecutor
    0.74
     ripe
    0.72
     precedent
    0.71
     supra
    0.70
    sov
    0.69
     Sever
    0.69
    Activations Density 1.726%

    No Known Activations