INDEX

Explanations

connective words indicating causality or logical sequence

oai_token-act-pair · gpt-3.5-turbo

the word "thus" used in various contexts to indicate a conclusion or result

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ten

-0.74

Kl

-0.71

 Children

-0.71

 Columb

-0.66

 Blue

-0.63

track

-0.63

 Price

-0.60

 Freddie

-0.60

AAA

-0.60

kick

-0.60

POSITIVE LOGITS

 guiActiveUn

0.87

wcs

0.81

 enthusi

0.78

 tiss

0.78

 philos

0.77

 dilig

0.77

ãĤ´ãĥ³

0.77

 conflic

0.77

è£ħ

0.76

 exha

0.75

Activations Density 0.008%

connective words indicating causality or logical sequence

the word "thus" used in various contexts to indicate a conclusion or result

No Comments

No Known Activations