INDEX

Explanations

mentions of possessive pronouns 'its' and 'his'

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

JUST

-0.67

****

-0.64

ILE

-0.61

anza

-0.59

cture

-0.58

Enabled

-0.58

nih

-0.58

***

-0.57

DEN

-0.57

Brow

-0.57

POSITIVE LOGITS

 attendant

1.22

 consequ

1.16

 accompanying

1.14

 ensuing

1.10

 associated

1.09

 subsequent

1.02

 surrounding

0.99

 corresponding

0.99

 resultant

0.98

 thereof

0.97

Activations Density 2.311%

mentions of possessive pronouns 'its' and 'his'

No Comments

No Known Activations