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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 10-RES_POST_32K-OAI
    6. 29471
    Prev
    Next
    INDEX
    Explanations

    financial and transactional language, particularly around payment and use

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    373
    +0.13
    0.6%
    316
    +0.13
    0.6%
    422
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    270
    +0.13
    0.02
    110
    +0.13
    0.05
    288
    +0.09
    0.08
    Negative Logits
    hap
    -0.81
    uga
    -0.71
    uously
    -0.71
    ractor
    -0.70
    poons
    -0.69
    netflix
    -0.69
    Rot
    -0.69
    imentary
    -0.69
    icularly
    -0.69
    Plex
    -0.68
    POSITIVE LOGITS
     attitude
    1.25
     mentality
    1.24
     proposition
    1.06
     syndrome
    1.05
     kind
    0.99
     approach
    0.97
     mindset
    0.92
     policy
    0.92
     mantra
    0.91
     clause
    0.90
    Activations Density 1.253%

    No Known Activations