© 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. 8116
    Prev
    Next
    INDEX
    Explanations

    punctuation marks, specifically apostrophes

    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
    314
    +0.15
    0.7%
    756
    +0.14
    0.7%
    102
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    102
    +0.15
    0.05
    152
    +0.14
    0.04
    314
    +0.12
    0.01
    Negative Logits
     brow
    -0.84
     cov
    -0.72
     throats
    -0.69
     synt
    -0.69
     synd
    -0.68
     print
    -0.68
     feature
    -0.68
     task
    -0.65
     scale
    -0.65
     share
    -0.65
    POSITIVE LOGITS
    ordinary
    1.07
    official
    1.05
    impro
    1.03
    independent
    1.01
    them
    0.99
    secret
    0.98
    liber
    0.98
    reve
    0.97
    til
    0.97
    behind
    0.96
    Activations Density 0.107%

    No Known Activations