© 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. 10-RES-JB
    11. 10188
    Prev
    Next
    INDEX
    Explanations

    phrases indicating the first in a series or comparison

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "one" and its variants, often indicating a sequence or listing

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    NRS
    -0.74
     ILCS
    -0.69
    lems
    -0.69
     unequ
    -0.62
    akh
    -0.62
    ÃįÃį
    -0.62
    Engineers
    -0.61
    WAR
    -0.58
    ARDS
    -0.58
    å½
    -0.57
    POSITIVE LOGITS
     consists
    1.26
     involves
    1.24
     consisted
    1.17
     relates
    1.16
     was
    1.13
     is
    1.09
     includes
    1.04
     lasted
    1.03
     contains
    1.00
     revolves
    0.99
    Activations Density 0.134%

    No Known Activations