© 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. 7553
    Prev
    Next
    INDEX
    Explanations

    titles and positions associated with leadership or authority roles

    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
    373
    +0.19
    0.9%
    436
    +0.12
    0.6%
    121
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    204
    +0.19
    0.06
    104
    +0.12
    0.06
    91
    +0.11
    0.06
    Negative Logits
    articles
    -0.74
     tracts
    -0.68
    Xi
    -0.64
    IRE
    -0.64
    rities
    -0.64
     verbs
    -0.63
     certificates
    -0.61
    Ts
    -0.61
     notions
    -0.61
    humans
    -0.60
    POSITIVE LOGITS
     overseeing
    0.98
     extraord
    0.87
     responsible
    0.79
     credited
    0.78
     hired
    0.78
     entrusted
    0.77
     tasked
    0.76
     who
    0.75
     whom
    0.75
     presiding
    0.74
    Activations Density 0.319%

    No Known Activations