© 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
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 4920
    Prev
    Next
    INDEX
    Explanations

    content related to digital media sources and links

    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
    629
    +0.10
    0.4%
    102
    +0.09
    0.4%
    50
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.10
    0.13
    314
    +0.09
    -0.02
    731
    +0.09
    0.05
    Negative Logits
     Advantage
    -0.89
     Ones
    -0.86
     Boards
    -0.83
     Opportun
    -0.82
     Tokens
    -0.82
     Regulations
    -0.82
     Boo
    -0.82
     Shades
    -0.81
     Offensive
    -0.81
     Anch
    -0.81
    POSITIVE LOGITS
    yahoo
    0.92
    internet
    0.89
    dragon
    0.84
    google
    0.83
     english
    0.83
    rw
    0.79
     google
    0.77
    united
    0.76
    json
    0.76
     typo
    0.76
    Activations Density 2.376%

    No Known Activations