INDEX

Explanations

the verb "be" in different contexts

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

rones

-0.72

 Notting

-0.71

Rab

-0.67

 Blitz

-0.64

 Desire

-0.63

erity

-0.62

 Mans

-0.61

-0.59

phabet

-0.59

riott

-0.58

POSITIVE LOGITS

 viewed

1.07

 traced

1.07

 seen

1.04

 construed

1.03

 easily

1.01

 likened

1.00

 considered

0.96

 forgiven

0.96

 undone

0.94

 accessed

0.92

Activations Density 0.348%

the verb "be" in different contexts

No Comments

No Known Activations