© 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. 8-RES_POST_32K-OAI
    11. 7941
    Prev
    Next
    INDEX
    Explanations

    phrases related to upgrades or modifications in technology

    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
    186
    +0.10
    0.4%
    194
    +0.09
    0.4%
    526
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.10
    0.10
    393
    +0.09
    0.10
    167
    +0.08
    0.00
    Negative Logits
    gers
    -0.69
     interactions
    -0.67
    akia
    -0.65
     shots
    -0.64
     thefts
    -0.62
     dealings
    -0.62
     finances
    -0.62
    ombs
    -0.60
     struggle
    -0.60
     abortions
    -0.60
    POSITIVE LOGITS
     appropriate
    1.03
    appropriate
    1.02
     correct
    0.97
     nearest
    0.91
     specified
    0.91
    specified
    0.87
     desired
    0.85
    approved
    0.85
     whichever
    0.85
    correct
    0.84
    Activations Density 1.106%

    No Known Activations