INDEX

Explanations

detailed descriptions or commentary on situations

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

iership

-0.84

ullivan

-0.80

ourses

-0.77

yr

-0.75

yss

-0.71

ovies

-0.71

ð

-0.71

atform

-0.71

ayn

-0.71

osponsors

-0.71

POSITIVE LOGITS

bit

1.27

 peek

0.92

 prick

0.85

BIT

0.81

Bit

0.80

 chuckle

0.78

 glimpse

0.76

dab

0.74

 tease

0.74

tad

0.74

Activations Density 1.801%

detailed descriptions or commentary on situations

No Comments

No Known Activations