INDEX

Explanations

emotional responses and reactions

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

 livest

-0.67

amins

-0.65

Ranked

-0.65

tein

-0.64

amation

-0.61

 landfill

-0.61

 wiki

-0.59

lay

-0.59

ublic

-0.58

ictionary

-0.58

POSITIVE LOGITS

ingly

1.04

 ABOUT

0.84

 about

0.81

NESS

0.76

wart

0.75

whel

0.73

der

0.73

dy

0.71

lessly

0.70

fully

0.70

Activations Density 3.813%

emotional responses and reactions

No Comments

No Known Activations