© 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. 8118
    Prev
    Next
    INDEX
    Explanations

    references to death and its impact on groups or communities

    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
    17
    +0.11
    0.5%
    222
    +0.11
    0.5%
    114
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    762
    +0.11
    0.12
    502
    +0.11
    0.06
    718
    +0.10
    0.09
    Negative Logits
    omatic
    -0.68
     bestselling
    -0.66
    iasis
    -0.64
    otype
    -0.62
    ital
    -0.61
     educator
    -0.61
     Effective
    -0.61
     Fant
    -0.60
     Career
    -0.60
    quote
    -0.60
    POSITIVE LOGITS
     gathered
    1.25
     scattered
    1.04
     disembark
    1.01
     assembled
    1.01
     clustered
    1.00
     congreg
    0.99
     evacuated
    0.97
    folk
    0.96
     piled
    0.96
    mates
    0.92
    Activations Density 1.501%

    No Known Activations