INDEX

Explanations

phrases that describe an event or situation

oai_token-act-pair · gpt-3.5-turbo

the repeated use of the word "was."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Current

-0.75

 IMAGES

-0.75

has

-0.72

onic

-0.70

arta

-0.68

continue

-0.67

LAT

-0.66

 Submit

-0.66

Make

-0.63

 Extend

-0.62

POSITIVE LOGITS

 unclear

1.21

 raining

1.16

 inevitable

0.97

 impossible

0.97

 rumored

0.92

 easier

0.90

 obvious

0.90

 easy

0.89

 fortunate

0.87

 evident

0.86

Activations Density 0.095%

phrases that describe an event or situation

the repeated use of the word "was."

No Comments

No Known Activations