INDEX

Explanations

quantitative information or statistics mentioned in a text

oai_token-act-pair · gpt-3.5-turbo

percentages and statistics related to public opinion and surveys

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

 laundering

-0.80

assetsadobe

-0.79

 scal

-0.77

adr

-0.74

 vectors

-0.73

 backups

-0.71

 graft

-0.70

 escapes

-0.70

route

-0.68

 Kafka

-0.67

POSITIVE LOGITS

 polled

1.22

 respondents

1.20

 surveyed

1.15

 Gallup

1.05

 undecided

1.05

 majorities

1.03

 respondent

1.02

Poll

0.96

 survey

0.95

 Opinion

0.94

Activations Density 0.656%

quantitative information or statistics mentioned in a text

percentages and statistics related to public opinion and surveys

No Comments

No Known Activations