INDEX

Explanations

information related to military conflicts and investigations

oai_token-act-pair · gpt-3.5-turbo

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 describ

-0.92

 neighb

-0.81

 mosqu

-0.78

 suspic

-0.75

 metic

-0.74

 destro

-0.74

 earthqu

-0.73

 petrol

-0.72

 spitting

-0.71

 sulph

-0.71

POSITIVE LOGITS

Advertisement

1.25

Advertisements

1.22

âĨĳ

1.20

RAW

1.15

Trivia

1.14

Discussion

1.14

Â©

1.13

WHERE

1.12

References

1.12

Â·

1.11

Activations Density 0.234%

information related to military conflicts and investigations

No Known Activations