INDEX

Explanations

phrases related to global or significant entities and events

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "world" or "global."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

veyard

-0.91

»Ĵ

-0.81

ippi

-0.80

sbm

-0.74

imon

-0.72

edin

-0.70

 COUNTY

-0.65

Mn

-0.65

ayers

-0.64

ividual

-0.63

POSITIVE LOGITS

 renowned

1.08

wide

1.03

 renown

0.96

 premiere

0.93

 championships

0.82

 wide

0.81

Wide

0.81

 championship

0.80

 heritage

0.80

liness

0.80

Activations Density 0.071%

phrases related to global or significant entities and events

references to the concept of "world" or "global."

No Comments

No Known Activations