INDEX

Explanations

the word "extra" and related concepts such as extraordinary, additional, or beyond normal

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

theless

-0.87

\\\\\\\\

-0.71

better

-0.70

geries

-0.67

 CrossRef

-0.66

Hes

-0.65

 Cosponsors

-0.65

 Associated

-0.63

eer

-0.63

bleacher

-0.62

POSITIVE LOGITS

judicial

1.42

ordinary

1.40

verted

1.04

veh

1.02

ord

0.95

verts

0.87

ngth

0.86

 virgin

0.86

vas

0.84

 goodies

0.83

Activations Density 0.023%

the word "extra" and related concepts such as extraordinary, additional, or beyond normal

No Comments

No Known Activations