INDEX

Explanations

conjunctions and transition phrases, particularly 'but'

oai_token-act-pair · gpt-3.5-turbo

the conjunction "but" and its variants to contrast or qualify statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

rill

-0.86

Osw

-0.72

à¥

-0.69

onne

-0.67

acca

-0.66

justice

-0.66

umat

-0.64

gd

-0.64

ilar

-0.63

ISH

-0.63

POSITIVE LOGITS

 hesitated

1.05

 lacks

1.01

 lacked

0.93

 alas

0.92

 nowhere

0.91

 declined

0.90

 none

0.89

 hasn

0.87

 denies

0.86

 fails

0.84

Activations Density 0.421%

conjunctions and transition phrases, particularly 'but'

the conjunction "but" and its variants to contrast or qualify statements

No Comments

No Known Activations