© 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. 8444
    Prev
    Next
    INDEX
    Explanations

    references to retinal or associated technology

    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
    0
    +0.12
    0.5%
    503
    +0.11
    0.5%
    502
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    503
    +0.12
    0.01
    651
    +0.11
    0.01
    232
    +0.09
    0.01
    Negative Logits
    WAYS
    -0.85
    theless
    -0.81
    姫
    -0.78
    ãģ®éŃĶ
    -0.74
    )=(
    -0.72
    SHIP
    -0.69
     Awakens
    -0.68
    ï¸
    -0.67
    lihood
    -0.64
    tenance
    -0.63
    POSITIVE LOGITS
    ainer
    1.17
    amping
    1.10
    raction
    1.03
    ribut
    1.02
    raining
    1.01
    ourn
    1.00
    amped
    0.96
    renched
    0.94
    ailed
    0.92
    reating
    0.91
    Activations Density 0.021%

    No Known Activations