INDEX

Explanations

information related to statistics, research reports, and international organizations

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

abases

-0.61

iencies

-0.60

agame

-0.59

aldehyde

-0.58

utherford

-0.58

ictional

-0.58

 chlorine

-0.55

aples

-0.55

geries

-0.54

ailability

-0.54

POSITIVE LOGITS

 said

1.13

 says

1.03

 told

0.97

 wrote

0.96

 Says

0.92

said

0.92

 tells

0.90

 explained

0.90

 notwithstanding

0.87

 explains

0.87

Activations Density 0.404%

information related to statistics, research reports, and international organizations

No Comments

No Known Activations