INDEX

Explanations

texts related to scientific experiments

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the term "experiment."

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

roots

-0.66

headers

-0.65

doms

-0.63

clinton

-0.61

cut

-0.61

veland

-0.59

 crisp

-0.58

CHAPTER

-0.56

othy

-0.56

yy

-0.56

POSITIVE LOGITS

imental

1.03

ally

1.03

iments

0.99

 Experiment

0.87

 experiment

0.85

iment

0.80

ually

0.79

ivity

0.76

ively

0.75

onom

0.74

Activations Density 0.035%

texts related to scientific experiments

repeated mentions of the term "experiment."

No Comments

No Known Activations