© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 5396
    Prev
    Next
    INDEX
    Explanations

    references to political events and legal matters

    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
    183
    +0.13
    0.6%
    69
    +0.12
    0.5%
    756
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.13
    0.14
    160
    +0.12
    0.14
    442
    +0.10
    0.13
    Negative Logits
    ultimate
    -0.80
     eq
    -0.76
    course
    -0.75
    alpha
    -0.73
    epad
    -0.71
    lvl
    -0.70
    prototype
    -0.69
    pg
    -0.68
     scales
    -0.68
    beta
    -0.67
    POSITIVE LOGITS
    WASHINGTON
    1.28
    ASHINGTON
    1.02
    SAN
    0.97
    LOS
    0.95
    Enlarge
    0.94
    Amid
    0.89
    window
    0.88
    Officials
    0.88
     UNCLASSIFIED
    0.87
    Updated
    0.86
    Activations Density 0.730%

    No Known Activations