INDEX

Explanations

declarative statements leading to some sort of conclusion

oai_token-act-pair · gpt-3.5-turbo

statements that assert knowledge or understanding about a topic

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 concess

-0.61

+.

-0.60

 hemor

-0.59

 frequ

-0.59

 sweats

-0.57

 heels

-0.56

 vain

-0.56

 downed

-0.54

 futile

-0.54

 classics

-0.53

POSITIVE LOGITS

 varies

0.89

 differs

0.85

 depends

0.84

is

0.83

 illustrates

0.81

 involves

0.77

 boils

0.75

SPONSORED

0.73

fold

0.71

 differed

0.71

Activations Density 0.169%

declarative statements leading to some sort of conclusion

statements that assert knowledge or understanding about a topic

No Comments

No Known Activations