INDEX

Explanations

terms related to regulations or restrictions, particularly those with a negative connotation such as 'dumping' and 'fraud'

oai_token-act-pair · gpt-3.5-turbo

terms related to anti-dumping measures and regulations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Abbey

-0.73

 Ples

-0.71

dq

-0.70

 belts

-0.65

 McKenna

-0.65

 Mock

-0.64

 Span

-0.63

ĸļ

-0.63

WD

-0.62

 Shib

-0.62

POSITIVE LOGITS

irt

0.76

amacare

0.76

anc

0.75

IR

0.74

uphem

0.74

roleum

0.73

NC

0.73

otation

0.73

elfare

0.73

ox

0.73

Activations Density 0.120%

terms related to regulations or restrictions, particularly those with a negative connotation such as 'dumping' and 'fraud'

terms related to anti-dumping measures and regulations

No Comments

No Known Activations