INDEX

Explanations

words related to strength or forceful attributes

oai_token-act-pair · gpt-3.5-turbo

references to intensity levels in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

vous

-0.81

nee

-0.73

BY

-0.73

OTA

-0.72

via

-0.69

unes

-0.69

olla

-0.68

rote

-0.68

ãĥīãĥ©

-0.67

 Passage

-0.67

POSITIVE LOGITS

 intensity

1.20

intensity

0.94

terness

0.88

 tremend

0.87

 è£ıè¦ļéĨĴ

0.83

 proble

0.80

mble

0.80

 interval

0.78

 urgency

0.78

 intens

0.78

Activations Density 0.009%

words related to strength or forceful attributes

references to intensity levels in various contexts

No Comments

No Known Activations