INDEX

Explanations

conjunctions, specifically the word "but" with emphasis

oai_token-act-pair · gpt-3.5-turbo

the conjunction "but" indicating contrast or exception in statements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

velt

-0.74

uto

-0.61

ocene

-0.60

ento

-0.60

eah

-0.60

 Combine

-0.60

mop

-0.58

 Peninsula

-0.58

resa

-0.58

ogical

-0.57

POSITIVE LOGITS

tons

1.13

chery

1.06

chers

0.96

 nevertheless

0.90

tered

0.90

 nonetheless

0.86

lers

0.86

ler

0.85

 alas

0.81

kus

0.80

Activations Density 0.096%

conjunctions, specifically the word "but" with emphasis

the conjunction "but" indicating contrast or exception in statements

No Comments

No Known Activations