INDEX

Explanations

verbs associated with spreading or revealing 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_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

BOOK

-0.73

BALL

-0.67

GAME

-0.65

 Samar

-0.63

Faw

-0.63

 Butt

-0.62

fulness

-0.61

Words

-0.60

Io

-0.60

LCS

-0.60

POSITIVE LOGITS

ated

2.16

ating

2.12

ation

1.94

ate

1.83

ates

1.79

ators

1.76

ator

1.73

ations

1.70

ative

1.52

ATED

1.50

Activations Density 0.072%

verbs associated with spreading or revealing information

No Comments

No Known Activations