© 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 EvalsExportsinterp-engineNEWAPI 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. 8369
    Prev
    Next
    INDEX
    Explanations

    phrases that convey confirmation or support for previous findings or sentiments

    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
    452
    +0.12
    0.5%
    71
    +0.10
    0.5%
    689
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.12
    0.11
    608
    +0.10
    0.10
    452
    +0.10
    0.04
    Negative Logits
    abilia
    -0.75
    illion
    -0.72
    cles
    -0.64
     Became
    -0.64
    abouts
    -0.63
     sucks
    -0.63
     Leilan
    -0.62
     fractions
    -0.60
    boxing
    -0.60
    iless
    -0.58
    POSITIVE LOGITS
     trend
    1.01
     earlier
    0.96
     longstanding
    0.95
     findings
    0.86
     previous
    0.85
     observation
    0.81
     sentiment
    0.81
     sentiments
    0.80
     recommendation
    0.79
     recommendations
    0.79
    Activations Density 1.563%

    No Known Activations