© 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. 3-RES-JB
    11. 11348
    Prev
    Next
    INDEX
    Explanations

    words related to virtues or positive qualities, particularly emphasizing characteristics like 'grace'

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "grace" and its variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ALK
    -0.76
    RAW
    -0.74
    WD
    -0.65
    ãĥĥãĥī
    -0.64
    OOL
    -0.63
    RAL
    -0.61
    ective
    -0.60
    irie
    -0.59
     iodine
    -0.59
    OTE
    -0.58
    POSITIVE LOGITS
    fully
    1.20
    fulness
    1.05
     grace
    0.90
    ibility
    0.89
    ful
    0.89
    ously
    0.84
    aciously
    0.83
    iffe
    0.82
    combe
    0.81
    ifully
    0.80
    Activations Density 0.009%

    No Known Activations