INDEX

Explanations

statements related to survey results or statistical data

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

ĪĴ

-0.75

 delicate

-0.72

 transition

-0.71

 reversible

-0.70

 seamless

-0.69

inge

-0.69

 handlers

-0.69

 suspension

-0.69

 discreet

-0.69

 speedy

-0.69

POSITIVE LOGITS

 Gallup

1.42

 According

1.29

Percent

1.20

Studies

1.19

Pew

1.17

 Surve

1.17

Statistics

1.17

According

1.15

 Median

1.13

according

1.13

Activations Density 0.565%

statements related to survey results or statistical data

No Comments

No Known Activations