© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 8850
    Prev
    Next
    INDEX
    Explanations

    phrases involving claims and accusations

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    544
    +0.11
    0.5%
    381
    +0.10
    0.5%
    53
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    528
    +0.11
    0.06
    393
    +0.10
    0.10
    566
    +0.09
    0.05
    Negative Logits
    pb
    -0.76
    izu
    -0.76
    olson
    -0.74
     doubtless
    -0.70
     mindset
    -0.70
    course
    -0.67
    meier
    -0.66
     dilemma
    -0.66
     certainly
    -0.65
     choices
    -0.64
    POSITIVE LOGITS
     cures
    0.90
     cured
    0.83
     aven
    0.73
     cure
    0.73
     mythical
    0.72
     paranormal
    0.70
     revenge
    0.69
     innocence
    0.69
     witchcraft
    0.69
     erad
    0.68
    Activations Density 1.794%

    No Known Activations