INDEX

Explanations

the word "all" appearing in the text

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aminer

-0.68

inth

-0.64

ritic

-0.64

SHIP

-0.63

assembly

-0.61

 Kamp

-0.61

dl

-0.61

ombat

-0.60

lav

-0.59

olulu

-0.58

POSITIVE LOGITS

ocating

1.13

iances

1.11

uding

1.05

udes

1.02

igators

1.01

iance

1.00

usions

0.96

oys

0.95

 bets

0.91

 sorts

0.91

Activations Density 0.079%

the word "all" appearing in the text

No Comments

No Known Activations