INDEX

Explanations

mentions of large entities or quantities

oai_token-act-pair · gpt-3.5-turbo

instances of the word "large" and variations of it in the context of size-related discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

çīĪ

-0.83

gnu

-0.74

emy

-0.72

CB

-0.72

emis

-0.69

forth

-0.69

ople

-0.68

onto

-0.67

Reviewed

-0.67

vous

-0.65

POSITIVE LOGITS

 intestine

1.15

 sized

1.08

 scale

1.05

 quantities

1.03

 amounts

1.00

(>

0.98

 denomination

0.93

 enough

0.85

 proport

0.84

 portions

0.82

Activations Density 0.034%

mentions of large entities or quantities

instances of the word "large" and variations of it in the context of size-related discussions

No Comments

No Known Activations