INDEX

Explanations

the word 'of'

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

Not in Any Lists

No Comments

Negative Logits

ynes

-0.64

 partName

-0.62

 reperto

-0.62

zer

-0.60

agher

-0.60

CPU

-0.59

Ô

-0.58

Angelo

-0.56

opus

-0.56

bet

-0.56

POSITIVE LOGITS

 course

1.81

course

1.61

 sorts

1.42

icial

1.16

 Course

1.16

 COUR

1.06

ortunately

1.02

 theirs

0.92

 ours

0.92

Course

0.88

Activations Density 1.473%

the word 'of'

No Comments

No Known Activations