© 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. 5015
    Prev
    Next
    INDEX
    Explanations

    symbols and characters commonly used in programming or mathematical contexts

    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
    314
    +0.12
    0.5%
    497
    +0.10
    0.5%
    485
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    675
    +0.12
    0.21
    650
    +0.10
    0.17
    314
    +0.10
    0.04
    Negative Logits
     Flavoring
    -0.74
     agre
    -0.66
     Alic
    -0.62
     Tot
    -0.61
     AGA
    -0.61
     lawy
    -0.60
     nerv
    -0.58
     lapt
    -0.57
     ACPI
    -0.57
    acles
    -0.56
    POSITIVE LOGITS
    âĸł
    0.96
    his
    0.80
    Shar
    0.78
    stellar
    0.77
    rain
    0.75
    none
    0.75
    present
    0.72
    eng
    0.69
    famous
    0.69
    split
    0.69
    Activations Density 1.390%

    No Known Activations