INDEX

Explanations

phrases related to controversy or conflicts

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

imester

-0.95

abase

-0.92

mediate

-0.85

flows

-0.80

cosystem

-0.80

 RELEASE

-0.80

esters

-0.79

Policy

-0.79

alysis

-0.78

 attrition

-0.78

POSITIVE LOGITS

 swast

1.45

 likeness

1.28

 logo

1.20

 logos

1.20

 initials

1.18

 adorned

1.14

 silhou

1.12

 emoji

1.11

 grinning

1.10

 smiling

1.10

Activations Density 6.218%

phrases related to controversy or conflicts

No Comments

No Known Activations