INDEX

Explanations

instances of the word "explained" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

luster

-0.71

ificial

-0.70

 ILCS

-0.70

illet

-0.70

 Sabha

-0.70

abases

-0.69

sembly

-0.68

Pont

-0.66

venge

-0.66

isol

-0.66

POSITIVE LOGITS

why

1.20

WHY

0.97

how

0.97

why

0.85

 succinct

0.85

 concise

0.78

 bluntly

0.76

how

0.76

 plainly

0.75

WER

0.75

Activations Density 0.029%

instances of the word "explained" in various contexts

No Comments

No Known Activations