INDEX

Explanations

news-related phrases

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

urat

-0.70

 expressive

-0.70

 discriminate

-0.65

otrop

-0.65

 blank

-0.65

cific

-0.63

 independ

-0.60

inav

-0.60

vag

-0.60

 dunno

-0.60

POSITIVE LOGITS

 Emails

1.04

Desk

0.89

 Coverage

0.80

 Desk

0.79

urrency

0.75

CNN

0.71

circ

0.69

 Latest

0.69

 Reporting

0.68

News

0.68

Activations Density 0.024%

news-related phrases

No Comments

No Known Activations

news-related phrases

No Comments

No Known Activations