INDEX

Explanations

phrases related to cause and effect

oai_token-act-pair · gpt-3.5-turbo

conjunctions that connect ideas or phrases, specifically the word "and."

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

uta

-0.59

Â¶

-0.55

:(

-0.55

:[

-0.53

CNN

-0.52

DX

-0.51

NULL

-0.51

 Powered

-0.50

Ĥİ

-0.50

BAT

-0.50

POSITIVE LOGITS

 consequently

1.01

 thus

0.99

 hence

0.96

 therefore

0.95

 thereby

0.94

 then

0.83

 vice

0.78

 thence

0.78

 secondly

0.78

 possibly

0.75

Activations Density 0.688%

phrases related to cause and effect

conjunctions that connect ideas or phrases, specifically the word "and."

No Comments

No Known Activations