© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 9108
    Prev
    Next
    INDEX
    Explanations

    terms related to perception and cognitive processes

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    194
    +0.10
    0.5%
    125
    +0.10
    0.5%
    678
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.10
    0.12
    738
    +0.10
    0.07
    678
    +0.09
    0.10
    Negative Logits
     Sov
    -0.80
     CoC
    -0.75
    ranch
    -0.72
    Sov
    -0.72
    eworks
    -0.71
     rebuilt
    -0.71
     Asylum
    -0.70
     disbanded
    -0.69
     corps
    -0.69
     Prohibition
    -0.68
    POSITIVE LOGITS
     stimuli
    1.50
     cues
    1.35
     stimulus
    1.16
     cue
    1.14
     stimulation
    1.04
     sensations
    1.03
     sensory
    1.03
     signals
    0.97
     sensation
    0.96
     arous
    0.92
    Activations Density 4.224%

    No Known Activations