INDEX

Explanations

superlative form adjectives that indicate the highest degree of a quality or rank

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "most" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

æ©

-0.76

wagen

-0.75

ategories

-0.71

raltar

-0.70

kas

-0.68

oak

-0.68

EStream

-0.68

allery

-0.68

too

-0.67

IDS

-0.67

POSITIVE LOGITS

 important

1.14

 efficient

1.13

 effective

1.01

 obvious

1.00

 expensive

0.99

 probable

0.98

 accurate

0.98

 desirable

0.98

 optimal

0.97

 egregious

0.96

Activations Density 0.056%

superlative form adjectives that indicate the highest degree of a quality or rank

references to the concept of "most" in various contexts

No Comments

No Known Activations