INDEX

Explanations

instructions or steps denoted by the word "should."

oai_token-act-pair · gpt-3.5-turbo

the word "should" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered 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

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

atile

-0.68

 Trails

-0.64

locked

-0.63

 Puzzle

-0.63

 Kore

-0.62

reality

-0.62

stood

-0.61

ils

-0.60

 Opposition

-0.59

GGGGGGGG

-0.58

POSITIVE LOGITS

be

0.97

ered

0.92

 ideally

0.91

 definitely

0.85

 probably

0.83

 preferably

0.82

 capitalize

0.80

ering

0.79

 certainly

0.78

nt

0.78

Activations Density 0.057%

instructions or steps denoted by the word "should."

the word "should" in various contexts

No Comments

No Known Activations