© 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
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 4290
    Prev
    Next
    INDEX
    Explanations

    instructions related to technology and software usage

    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
    519
    +0.12
    0.5%
    373
    +0.11
    0.5%
    737
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.12
    0.15
    435
    +0.11
    0.10
    698
    +0.11
    0.10
    Negative Logits
    bidden
    -0.63
     coron
    -0.60
     overshadowed
    -0.57
    was
    -0.56
    CVE
    -0.56
    ihadi
    -0.55
    "]=>
    -0.55
    ilts
    -0.55
    oln
    -0.55
    agged
    -0.54
    POSITIVE LOGITS
     yourself
    1.32
     preferably
    1.18
     yourselves
    1.08
     ASAP
    1.05
     Yourself
    1.04
     wisely
    1.01
     BEFORE
    0.91
     THEN
    0.89
     Then
    0.86
     anytime
    0.85
    Activations Density 2.372%

    No Known Activations