© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 2740
    Prev
    Next
    INDEX
    Explanations

    words related to lightness or illumination

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "lighter" and related variations, possibly focusing on items or contexts that suggest lightness or subtlety

    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
    ral
    -0.66
     seal
    -0.65
     silicon
    -0.65
     plac
    -0.63
     location
    -0.62
     vacancies
    -0.62
     locations
    -0.62
     plat
    -0.60
     pin
    -0.60
     corrupt
    -0.59
    POSITIVE LOGITS
    ighter
    4.13
    IGHT
    3.48
    ighters
    3.41
    ights
    2.25
    IGHTS
    2.22
    ighty
    1.80
    ight
    1.79
    ighting
    1.68
    IGH
    1.68
    igh
    1.44
    Activations Density 0.018%

    No Known Activations