INDEX

Explanations

instances where a causal relationship is being discussed or indicated

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Doctrine

-0.69

 Thoughts

-0.63

ONE

-0.61

Eye

-0.59

 boarding

-0.59

 Gallery

-0.59

 guessing

-0.58

————

-0.56

 Basics

-0.56

 Planning

-0.56

POSITIVE LOGITS

 resembles

1.14

 extends

1.08

 accompanies

1.07

 includes

1.04

 spans

1.03

 incorporates

1.03

 satisfies

1.00

 encompasses

1.00

 promotes

1.00

 enables

1.00

Activations Density 0.678%

instances where a causal relationship is being discussed or indicated

No Comments

No Known Activations