INDEX

Explanations

questions or queries asking for opinions or decisions

oai_token-act-pair · gpt-3.5-turbo

conditional or hypothetical questions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

natureconservancy

-0.78

ãĤĬ

-0.72

SPONSORED

-0.70

ãģĮ

-0.69

ovember

-0.64

ãģ«

-0.63

perty

-0.62

 awoke

-0.60

 scrimmage

-0.57

ufact

-0.56

POSITIVE LOGITS

1.19

olated

1.01

 anyone

1.00

 anybody

0.98

olation

0.88

abella

0.85

olate

0.85

zens

0.83

 there

0.81

 Anyone

0.73

Activations Density 0.072%

questions or queries asking for opinions or decisions

conditional or hypothetical questions

No Comments

No Known Activations