INDEX

Explanations

comparisons indicating improvement or superiority

oai_token-act-pair · gpt-3.5-turbo

instances of the word "better."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

iasco

-0.70

lies

-0.69

 Contents

-0.63

endant

-0.61

ital

-0.60

lig

-0.60

opsis

-0.60

aline

-0.59

achusetts

-0.59

achi

-0.59

POSITIVE LOGITS

 better

3.42

better

2.85

 Better

2.07

 worse

2.06

Better

2.06

 nicer

2.05

 smarter

1.95

 safer

1.94

 stronger

1.81

 wiser

1.76

Activations Density 0.032%

comparisons indicating improvement or superiority

instances of the word "better."

No Comments

No Known Activations