INDEX

Explanations

phrases related to providing assistance or aid to specific countries or regions

oai_token-act-pair · gpt-3.5-turbo

phrases related to foreign aid and military support

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ebook

-0.85

prints

-0.81

ãĤ¦ãĤ¹

-0.75

glass

-0.69

bryce

-0.67

concept

-0.67

levision

-0.67

workshop

-0.65

attribute

-0.64

ider

-0.64

POSITIVE LOGITS

 embattled

1.13

 needy

1.11

 victims

1.10

 rebels

1.06

 vetted

1.00

 persecuted

0.98

 stricken

0.96

 impoverished

0.96

 survivors

0.96

 Syrians

0.95

Activations Density 0.232%

phrases related to providing assistance or aid to specific countries or regions

phrases related to foreign aid and military support

No Known Activations