INDEX

Explanations

phrases related to rules, responsibility, and consequences

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

anka

-0.63

OTOS

-0.63

htaking

-0.61

 Penny

-0.60

tyard

-0.58

atl

-0.57

 Vance

-0.57

 Hass

-0.56

 Moroc

-0.56

Favorite

-0.56

POSITIVE LOGITS

 automatically

1.03

 surely

1.02

 forfeit

0.94

 likely

0.93

 chances

0.86

 probably

0.85

 inevitably

0.83

 immediately

0.81

 invariably

0.80

 nevertheless

0.80

Activations Density 3.351%

phrases related to rules, responsibility, and consequences

No Comments

No Known Activations