INDEX

Explanations

phrases used to provide clarification or emphasize a point

oai_token-act-pair · gpt-3.5-turbo

the expression of disbelief or emphasis in statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

"},"

-0.72

onding

-0.65

ittered

-0.64

-+-+

-0.63

 corrid

-0.63

affe

-0.61

iets

-0.59

appers

-0.58

icipated

-0.57

iated

-0.56

POSITIVE LOGITS

 seriously

1.08

 honestly

1.00

 REALLY

0.99

 yeah

0.98

yea

0.90

 LOOK

0.86

 literally

0.84

 Seriously

0.83

wow

0.82

 sure

0.81

Activations Density 0.022%

phrases used to provide clarification or emphasize a point

the expression of disbelief or emphasis in statements

No Comments

No Known Activations