INDEX

Explanations

words related to negative occurrences or challenges faced by individuals or organizations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tein

-0.86

 Pastebin

-0.67

 prem

-0.61

 vantage

-0.61

ourage

-0.61

 principals

-0.60

 Essex

-0.60

XV

-0.58

 township

-0.58

 unification

-0.55

POSITIVE LOGITS

ãĤ¦ãĤ¹

0.90

ãĤ°

0.87

ãĥ¼ãĤ¯

0.84

gaard

0.83

ãĤ¼

0.79

ãĤ¡

0.79

 aback

0.78

ãĥīãĥ©

0.77

Ð¼

0.77

ÙĪ

0.77

Activations Density 0.086%

words related to negative occurrences or challenges faced by individuals or organizations

No Comments

No Known Activations