INDEX

Explanations

phrases indicating speech or opinion

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

 ILCS

-0.74

xtap

-0.66

odore

-0.65

brance

-0.64

pees

-0.64

Cop

-0.63

zens

-0.63

isable

-0.63

\/\/

-0.63

FILE

-0.62

POSITIVE LOGITS

 bluntly

0.98

 sarcast

0.92

:"

0.92

"[

0.89

 goodbye

0.88

 aloud

0.87

 anecd

0.84

 "âĢ¦

0.83

:[

0.82

 succinct

0.81

Activations Density 0.268%

phrases indicating speech or opinion

No Comments

No Known Activations