INDEX

Explanations

words associated with assigning blame or responsibility

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ById

-0.69

Dur

-0.67

 pops

-0.66

 deductions

-0.62

 Advertisement

-0.62

Got

-0.61

 didnt

-0.61

ain

-0.60

 sources

-0.60

 reperto

-0.60

POSITIVE LOGITS

asted

1.35

asting

1.26

wered

1.09

asters

1.06

ying

1.06

pless

1.03

 blame

1.01

lled

1.00

othy

0.98

asty

0.97

Activations Density 0.075%

words associated with assigning blame or responsibility

No Comments

No Known Activations