© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 5404
    Prev
    Next
    INDEX
    Explanations

    references to the Freedom of Information Act and related transparency issues

    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
    140
    +0.10
    0.5%
    164
    +0.10
    0.5%
    286
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.04
    225
    +0.10
    0.03
    140
    +0.09
    0.03
    Negative Logits
    vic
    -0.70
    Interstitial
    -0.69
    nar
    -0.69
    Stars
    -0.68
    atomic
    -0.62
     nood
    -0.62
    gran
    -0.62
     mell
    -0.62
    Population
    -0.61
     totality
    -0.61
    POSITIVE LOGITS
     subpoen
    0.97
     disclosures
    0.96
     FOIA
    0.94
     subpoena
    0.92
     whistleblower
    0.89
    recy
    0.88
     requests
    0.83
    acious
    0.82
     revelations
    0.81
     documents
    0.79
    Activations Density 0.185%

    No Known Activations