© 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. 8858
    Prev
    Next
    INDEX
    Explanations

    negative constructions and limitations in statements

    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
    1
    +0.10
    0.5%
    340
    +0.10
    0.4%
    527
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    20
    +0.10
    0.07
    14
    +0.10
    0.05
    126
    +0.09
    0.06
    Negative Logits
    terness
    -0.97
    icism
    -0.96
    matter
    -0.95
    iliation
    -0.93
    leness
    -0.81
    analysis
    -0.80
    cation
    -0.79
    eness
    -0.78
    peed
    -0.77
    activity
    -0.76
    POSITIVE LOGITS
     exceptions
    0.97
     indications
    0.95
     few
    0.89
     occasions
    0.88
     constants
    0.85
     moments
    0.83
     teaspoons
    0.83
     instances
    0.81
     signs
    0.80
     reasons
    0.79
    Activations Density 0.332%

    No Known Activations