INDEX

Explanations

semi-colons representing a shift in thought or emphasis 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

orts

-0.76

Bas

-0.71

enium

-0.68

IDENT

-0.68

WAY

-0.67

ema

-0.67

din

-0.66

unda

-0.66

Accessory

-0.66

ene

-0.65

POSITIVE LOGITS

 namely

1.15

-)

1.08

 moreover

1.07

 furthermore

1.05

 hence

1.02

 alas

1.00

 alternatively

0.97

 indeed

0.94

 nevertheless

0.94

 however

0.91

Activations Density 0.038%

semi-colons representing a shift in thought or emphasis in the text

No Comments

No Known Activations