INDEX

Explanations

the word "read" in various contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "read" and its variations, indicating a focus on reading-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Goodman

-0.68

Zur

-0.67

xon

-0.66

ortium

-0.61

 Revenue

-0.59

Ø§ÙĦ

-0.59

 Michaels

-0.59

Ã±o

-0.57

anova

-0.56

Aus

-0.56

POSITIVE LOGITS

 aloud

1.27

just

1.24

ying

1.19

ied

1.14

dress

1.05

 comprehension

0.99

mitt

0.98

mission

0.96

ies

0.92

mitted

0.92

Activations Density 0.046%

the word "read" in various contexts

instances of the word "read" and its variations, indicating a focus on reading-related concepts

No Comments

No Known Activations