INDEX

Explanations

instances of the word "that" used to introduce an explanation or a defining clause

oai_token-act-pair · gpt-3.5-turbo

the recurring phrase "the fact that," indicating it is looking for statements presenting assertions or conclusions

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

OLOG

-0.68

mes

-0.68

ãĥ¼ãĥ³

-0.67

YC

-0.65

OHN

-0.64

Ble

-0.64

rup

-0.62

HO

-0.61

robe

-0.61

INAL

-0.61

POSITIVE LOGITS

 they

0.76

 pesky

0.69

 contradicts

0.68

 hindsight

0.67

chery

0.65

soever

0.65

zsche

0.65

 accompanies

0.63

 happened

0.63

we

0.62

Activations Density 0.086%

instances of the word "that" used to introduce an explanation or a defining clause

the recurring phrase "the fact that," indicating it is looking for statements presenting assertions or conclusions

No Comments

No Known Activations