INDEX

Explanations

words related to statistical analysis and results

oai_token-act-pair · gpt-3.5-turbo

quantitative expressions and significant results from research studies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

Edit

-0.78

quer

-0.77

wered

-0.75

soever

-0.72

advertising

-0.71

its

-0.68

 helicop

-0.67

 Provided

-0.67

agonist

-0.67

Ask

-0.66

POSITIVE LOGITS

 discrepancy

1.45

 similarity

1.40

 resemblance

1.34

 similarities

1.30

 correlation

1.24

 difference

1.23

 disparity

1.22

 discrepancies

1.18

 inconsistency

1.17

 divergence

1.16

Activations Density 0.343%

words related to statistical analysis and results

quantitative expressions and significant results from research studies

No Comments

No Known Activations