INDEX

Explanations

emotional or charged language related to societal issues and controversies

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

OPLE

-0.77

 briefs

-0.74

Tut

-0.71

 Madden

-0.68

 Blitz

-0.67

hetti

-0.66

 Ajax

-0.65

 slopes

-0.64

Ank

-0.64

MX

-0.63

POSITIVE LOGITS

balanced

1.25

ruly

1.25

assuming

1.23

earned

1.21

cles

1.20

readable

1.19

ifying

1.17

usual

1.14

availability

1.13

ipolar

1.13

Activations Density 0.309%

emotional or charged language related to societal issues and controversies

No Comments

No Known Activations