INDEX

Explanations

numerical figures presented with a specific format

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 barr

-0.91

 bonded

-0.90

 casc

-0.89

 sculpt

-0.88

 safegu

-0.87

 cooper

-0.86

 interven

-0.85

 stressed

-0.85

 tink

-0.85

hyp

-0.85

POSITIVE LOGITS

Average

2.02

Percent

2.00

Rank

1.94

Year

1.93

Total

1.84

Number

1.79

Hig

1.73

SELECT

1.72

Avg

1.71

Country

1.71

Activations Density 0.211%

numerical figures presented with a specific format

No Comments

No Known Activations