INDEX

Explanations

passive voice constructions

oai_token-act-pair · gpt-3.5-turbo

phrases related to legislative action or legal votes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ebus

-0.72

eur

-0.65

oler

-0.63

waukee

-0.63

drive

-0.63

Registered

-0.62

pipe

-0.62

WT

-0.61

oux

-0.61

enser

-0.60

POSITIVE LOGITS

 grouped

1.07

 categories

0.97

 combinations

0.96

:-

0.95

 nominees

0.94

 numbered

0.94

 finalists

0.93

 subsections

0.92

 listed

0.91

 names

0.91

Activations Density 1.316%

passive voice constructions

phrases related to legislative action or legal votes

No Comments

No Known Activations