INDEX

Explanations

words related to agreements or deals between entities

oai_token-act-pair · gpt-3.5-turbo

instances of the word "agree" along with its derivatives 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 @ 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

grain

-0.75

patch

-0.72

clips

-0.69

pointer

-0.68

 wastes

-0.66

ults

-0.65

 Detected

-0.64

worm

-0.64

worms

-0.62

ï¸ı

-0.61

POSITIVE LOGITS

 waive

1.06

 cooperate

1.04

 accept

1.03

 settle

0.97

 participate

0.96

 obey

0.95

 submit

0.95

 surrender

0.94

 abide

0.93

 undertake

0.89

Activations Density 0.062%

words related to agreements or deals between entities

instances of the word "agree" along with its derivatives in various contexts

No Comments

No Known Activations