© 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. 7757
    Prev
    Next
    INDEX
    Explanations

    phrases indicating possession or experience related to personal actions

    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
    518
    +0.11
    0.5%
    173
    +0.11
    0.5%
    75
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    566
    +0.11
    0.10
    183
    +0.11
    0.07
    344
    +0.10
    0.09
    Negative Logits
     broadly
    -0.74
     thereby
    -0.70
    atform
    -0.69
    enance
    -0.68
    NH
    -0.68
     secondly
    -0.68
     principally
    -0.68
     hereafter
    -0.67
    arger
    -0.66
    large
    -0.65
    POSITIVE LOGITS
     gotta
    1.03
     fuckin
    0.88
     fucking
    0.86
     gonna
    0.82
     goddamn
    0.78
     fucked
    0.74
     godd
    0.72
     Godd
    0.72
     ain
    0.70
     Ruin
    0.70
    Activations Density 0.396%

    No Known Activations