INDEX

Explanations

phrases that express contrast or opposition

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sappi

-1.15

 peculi

-1.14

 incess

-1.12

 alkoh

-1.10

 meis

-1.06

 franz

-1.06

 pecuni

-1.03

 nomine

-1.00

 inder

-1.00

 doman

-0.99

POSITIVE LOGITS

 fact

1.09

 indeed

1.07

 Indeed

1.02

 actually

1.01

indeed

0.96

 Actually

0.94

Indeed

0.91

actually

0.89

むしろ

0.83

Actually

0.79

Activations Density 0.267%

phrases that express contrast or opposition

No Comments

No Known Activations