INDEX

Explanations

percentages and numerical statistics in text, focusing on survey data and statistical information

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

 jerk

-0.62

elf

-0.61

 exile

-0.61

 histories

-0.60

 fame

-0.60

 bombard

-0.59

 marqu

-0.59

 besie

-0.59

 vandal

-0.59

Dos

-0.59

POSITIVE LOGITS

1.41

 percent

1.35

%)

1.21

%),

1.20

%;

1.19

%,

1.19

%.

1.19

%).

1.18

%"

1.14

%-

1.13

Activations Density 0.893%

percentages and numerical statistics in text, focusing on survey data and statistical information

No Comments

No Known Activations