© 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. 5606
    Prev
    Next
    INDEX
    Explanations

    references to seduction or seductive behavior

    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
    491
    +0.10
    0.5%
    423
    +0.10
    0.4%
    554
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.01
    201
    +0.10
    0.01
    367
    +0.09
    0.01
    Negative Logits
    SHIP
    -0.70
    ãĥ£
    -0.67
     Briggs
    -0.66
    Torrent
    -0.62
    BY
    -0.61
    pei
    -0.61
    Ble
    -0.61
     Union
    -0.60
    boards
    -0.58
    #$
    -0.57
    POSITIVE LOGITS
    entary
    1.36
    uctive
    1.33
    iments
    1.30
    uction
    1.20
    ucing
    1.15
    uct
    1.15
    uced
    1.15
    uce
    1.12
    iment
    1.08
    imentary
    1.06
    Activations Density 0.056%

    No Known Activations