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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 10-RES_POST_32K-OAI
    11. 26818
    Prev
    Next
    INDEX
    Explanations

    instances of transitions, changes, or shifts in various contexts

    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
    462
    +0.13
    0.6%
    218
    +0.11
    0.5%
    398
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.13
    0.11
    362
    +0.11
    0.08
    647
    +0.10
    0.05
    Negative Logits
    icious
    -0.68
     resemb
    -0.65
     misunderstand
    -0.63
     woes
    -0.62
     disappoint
    -0.62
     mistakes
    -0.62
     attest
    -0.61
     resemblance
    -0.60
    hoe
    -0.59
    iae
    -0.58
    POSITIVE LOGITS
     safer
    0.97
     cheaper
    0.90
     newer
    0.90
     another
    0.89
     healthier
    0.84
     cleaner
    0.83
     new
    0.81
    another
    0.80
     quieter
    0.73
     simpler
    0.68
    Activations Density 0.924%

    No Known Activations