INDEX

Explanations

codes and technical terms in a mixed context

oai_token-act-pair · gpt-3.5-turbo

foreign language or multilingual content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

angelo

-0.71

 specificity

-0.59

baum

-0.57

GROUP

-0.55

 cautiously

-0.55

arers

-0.53

hetically

-0.52

afety

-0.52

 bearer

-0.52

taker

-0.52

POSITIVE LOGITS

ĪĴ

0.93

orld

0.73

 Wonderland

0.66

enegger

0.60

à¦

0.60

worm

0.59

 trout

0.59

Textures

0.59

 Lith

0.58

cir

0.55

Activations Density 1.062%

codes and technical terms in a mixed context

foreign language or multilingual content

No Comments

No Known Activations