© 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. 26984
    Prev
    Next
    INDEX
    Explanations

    references to published works and their bibliographic details

    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
    +0.11
    0.5%
    6
    +0.11
    0.5%
    165
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    426
    +0.11
    0.05
    317
    +0.11
    0.07
    630
    +0.10
    0.07
    Negative Logits
     diapers
    -0.78
     resil
    -0.77
     cooled
    -0.74
    */(
    -0.73
     bonding
    -0.71
     curtains
    -0.70
     neighb
    -0.70
     setups
    -0.69
     carbs
    -0.69
     babies
    -0.69
    POSITIVE LOGITS
     Ibid
    1.77
     Cf
    1.27
     cf
    1.16
     ib
    1.07
    ibliography
    1.04
     See
    1.03
     footnote
    0.98
     Interview
    0.96
     quoting
    0.95
     reprinted
    0.95
    Activations Density 0.890%

    No Known Activations