INDEX

Explanations

the word "when" followed by a numerical value

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enegger

-0.74

kaya

-0.71

gur

-0.69

zzi

-0.68

yi

-0.66

feature

-0.65

edly

-0.62

hid

-0.61

hire

-0.61

chens

-0.61

POSITIVE LOGITS

soever

1.04

 exactly

0.95

abouts

0.79

ce

0.79

irlf

0.78

 they

0.73

 someone

0.71

 puberty

0.65

we

0.65

 faced

0.65

Activations Density 0.078%

the word "when" followed by a numerical value

No Comments

No Known Activations