INDEX

Explanations

adverbs that express an opinion or evaluation

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ories

-0.64

 aspiration

-0.62

SourceFile

-0.59

 Variant

-0.58

mates

-0.58

CLOSE

-0.57

(>

-0.57

nect

-0.56

MAL

-0.56

POSITIVE LOGITS

 anecd

0.73

 unlike

0.71

 according

0.67

 concedes

0.67

adle

0.67

âķĲâķĲ

0.66

 ignores

0.66

requires

0.64

has

0.64

 acknowledges

0.63

Activations Density 0.102%

adverbs that express an opinion or evaluation

No Comments

No Known Activations