INDEX

Explanations

instances where the word "it" is followed by another word or phrase

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate uncertainty or conditionality

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Et

-0.72

TT

-0.72

ands

-0.71

onement

-0.70

Ts

-0.69

IGN

-0.69

soDeliveryDate

-0.69

lette

-0.68

lets

-0.67

onds

-0.67

POSITIVE LOGITS

 technically

1.03

 disagree

0.99

 admittedly

0.98

 differed

0.96

 disagreed

0.93

 slight

0.92

 occasional

0.87

 initially

0.85

 differs

0.84

 concede

0.83

Activations Density 0.236%

instances where the word "it" is followed by another word or phrase

phrases that indicate uncertainty or conditionality

No Comments

No Known Activations