INDEX

Explanations

negations or negative expressions

oai_token-act-pair · gpt-3.5-turbo

negations or phrases indicating what is not the case

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

soType

-0.78

 Strategy

-0.65

è¦ļéĨĴ

-0.64

 Designs

-0.62

 Slam

-0.62

 Silk

-0.61

 Statements

-0.60

stakes

-0.60

 Strategies

-0.59

 Showdown

-0.59

POSITIVE LOGITS

icably

1.12

icable

1.02

 necessarily

1.01

 adequately

1.01

 bothered

0.99

 bother

0.99

existent

0.94

epad

0.92

hin

0.89

 bothering

0.87

Activations Density 0.446%

negations or negative expressions

negations or phrases indicating what is not the case

No Comments

No Known Activations