INDEX

Explanations

transitional words indicating a logical consequence or conclusion

oai_token-act-pair · gpt-3.5-turbo

conjunctions and transitional phrases that indicate conclusions or follow-ups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 neighbourhood

-0.66

 upgr

-0.60

igi

-0.59

duction

-0.58

 Rumble

-0.58

 neighborhood

-0.57

 smack

-0.57

 Caribbean

-0.56

 area

-0.55

 harbour

-0.55

POSITIVE LOGITS

fter

0.85

hester

0.83

resa

0.81

forward

0.81

hovah

0.79

apy

0.78

ccording

0.75

weet

0.71

noon

0.70

ommel

0.70

Activations Density 0.019%

transitional words indicating a logical consequence or conclusion

conjunctions and transitional phrases that indicate conclusions or follow-ups

No Comments

No Known Activations