INDEX

Explanations

comparisons indicating something is more significant or complex than initially perceived

oai_token-act-pair · gpt-3.5-turbo

comparative phrases that indicate something is "more than" a particular quality or quantity

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

taboola

-0.88

ãĤ´ãĥ³

-0.81

ella

-0.80

ECT

-0.79

ONS

-0.76

Winged

-0.76

âĹ¼

-0.75

ãĤ¨

-0.74

Constructed

-0.70

éŃĶ

-0.70

POSITIVE LOGITS

 mere

1.30

 merely

1.22

 simply

1.19

 just

1.14

 meets

1.13

 simple

0.93

 superficial

0.92

 JUST

0.87

 brute

0.83

 cosmetic

0.81

Activations Density 0.104%

comparisons indicating something is more significant or complex than initially perceived

comparative phrases that indicate something is "more than" a particular quality or quantity

No Comments

No Known Activations