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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 5619
    Prev
    Next
    INDEX
    Explanations

    descriptors related to clarity, communication, and articulateness in writing

    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
    634
    +0.09
    0.4%
    136
    +0.09
    0.4%
    710
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.09
    0.07
    159
    +0.09
    0.05
    737
    +0.09
    0.05
    Negative Logits
    ota
    -0.82
     Skydragon
    -0.81
     Shrine
    -0.72
    slave
    -0.72
    olen
    -0.72
     Alive
    -0.72
     Stores
    -0.71
    rha
    -0.68
     Takeru
    -0.66
    hops
    -0.65
    POSITIVE LOGITS
     critiques
    1.22
     critique
    1.19
     criticism
    1.18
     condemnation
    1.17
     criticisms
    1.14
     indictment
    1.13
     denounce
    1.09
     rebutt
    1.07
     speeches
    1.06
     statements
    1.04
    Activations Density 0.701%

    No Known Activations