© 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. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 15554
    Prev
    Next
    INDEX
    Explanations

    words related to marking or evaluation

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating notable events or evaluations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    witch
    -0.74
    hip
    -0.68
    ystem
    -0.65
     HI
    -0.64
    umo
    -0.64
    aghetti
    -0.62
     Podesta
    -0.61
    gger
    -0.61
    Resource
    -0.60
     Sov
    -0.60
    POSITIVE LOGITS
     marks
    4.18
     mark
    2.51
    marks
    2.31
     Marks
    2.30
     marking
    1.74
    mark
    1.69
     markings
    1.65
     marked
    1.61
    marked
    1.59
     scars
    1.39
    Activations Density 0.014%

    No Known Activations