INDEX

Explanations

questions starting with "Why"

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

ibaba

-0.69

semble

-0.67

 Roller

-0.66

izen

-0.66

ymph

-0.64

lator

-0.63

 consolation

-0.62

jri

-0.59

culosis

-0.58

mun

-0.58

POSITIVE LOGITS

why

1.28

why

1.11

WHY

1.09

abl

0.93

 bother

0.88

Why

0.87

Origin

0.80

Why

0.78

forth

0.72

 motives

0.69

Activations Density 3.896%

questions starting with "Why"

No Comments

No Known Activations