INDEX

Explanations

words related to controversy and criticism

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

 Occupations

-0.81

Construct

-0.66

UAL

-0.66

AGE

-0.64

reon

-0.64

 Gund

-0.63

uality

-0.63

Condition

-0.61

allic

-0.61

BOOK

-0.60

POSITIVE LOGITS

ossom

1.27

itzer

1.23

anche

1.21

ocking

1.12

adders

1.11

izzard

1.10

anco

1.10

anca

1.07

asts

1.02

acks

1.02

Activations Density 0.407%

words related to controversy and criticism

No Comments

No Known Activations