© 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. 4155
    Prev
    Next
    INDEX
    Explanations

    terms associated with negativity or criticism, particularly in a political or social context

    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
    426
    +0.11
    0.5%
    287
    +0.10
    0.4%
    648
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    331
    +0.11
    0.08
    616
    +0.10
    0.07
    314
    +0.09
    -0.01
    Negative Logits
     waivers
    -0.82
     certification
    -0.68
     Contracts
    -0.67
     designation
    -0.64
     RI
    -0.64
     Wolves
    -0.63
     doubles
    -0.62
     accuser
    -0.62
    ISA
    -0.61
     jails
    -0.60
    POSITIVE LOGITS
    ensical
    1.20
    erous
    1.08
    icative
    1.02
    entious
    1.00
    ceptive
    1.00
    ocative
    0.97
    isive
    0.97
    icky
    0.96
    igent
    0.96
    ulent
    0.94
    Activations Density 1.313%

    No Known Activations