INDEX

Explanations

quantitative expressions involving ordinal numbers

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

stress

-0.64

impl

-0.62

Impl

-0.61

hement

-0.61

NOW

-0.59

sym

-0.58

agic

-0.58

aceae

-0.58

 subsequ

-0.57

antis

-0.56

POSITIVE LOGITS

ifiable

1.22

ifications

1.00

shy

0.94

ices

0.79

IFIC

0.77

ifier

0.71

ifiers

0.69

 cents

0.68

iciary

0.68

Activations Density 0.071%

quantitative expressions involving ordinal numbers

No Comments

No Known Activations