INDEX

Explanations

references to different types of loss, including physical, emotional, and financial losses

oai_token-act-pair · gpt-3.5-turbo

terms related to loss, particularly in contexts of damage or casualties

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.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

ansky

-0.75

ECK

-0.71

ATT

-0.68

ENTS

-0.68

rouse

-0.64

 inventive

-0.62

":[{"

-0.60

chell

-0.59

âĸ¬âĸ¬

-0.59

andise

-0.59

POSITIVE LOGITS

 aversion

0.97

 incurred

0.95

iem

0.92

 Loss

0.83

 loss

0.79

luster

0.79

esville

0.79

 losses

0.75

 suffered

0.73

loss

0.73

Activations Density 0.023%

references to different types of loss, including physical, emotional, and financial losses

terms related to loss, particularly in contexts of damage or casualties

No Comments

No Known Activations