© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 1950
    Prev
    Next
    INDEX
    Explanations

    words related to causality or consequence

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "thus" indicating a causal or explanatory relationship

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    kick
    -0.70
    ten
    -0.68
     Kl
    -0.67
     Metro
    -0.66
     metro
    -0.64
     anten
    -0.61
     Kelvin
    -0.60
     Williamson
    -0.60
     Ones
    -0.58
     Lynn
    -0.58
    POSITIVE LOGITS
    forth
    1.19
    bered
    0.84
    hiba
    0.80
    mia
    0.79
    apers
    0.77
    apy
    0.77
    forward
    0.77
     far
    0.76
    mask
    0.76
     philos
    0.75
    Activations Density 0.015%

    No Known Activations