INDEX

Explanations

information related to volumes, years, and specific institutions

oai_token-act-pair · gpt-3.5-turbo

references to various volumes, likely in the context of publications or series

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

-0.67

ãĥķãĤ©

-0.65

ÄŁ

-0.57

ellen

-0.57

kl

-0.56

 Amit

-0.54

packing

-0.53

Ain

-0.53

rament

-0.52

 Bethlehem

-0.52

POSITIVE LOGITS

icut

0.71

 Wraith

0.67

 targ

0.65

ucer

0.60

endif

0.60

 Viper

0.59

ucket

0.58

¢

0.58

pty

0.58

uls

0.57

Activations Density 0.470%

information related to volumes, years, and specific institutions

references to various volumes, likely in the context of publications or series

No Comments

No Known Activations