INDEX

Explanations

phrases related to things being up-to-date or upcoming

oai_token-act-pair · gpt-3.5-turbo

phrases indicating trends or changes over time

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

eri

-0.68

 shorth

-0.68

èĢħ

-0.67

Äĩ

-0.65

hiba

-0.65

Anne

-0.65

WARNING

-0.64

Mah

-0.62

ucci

-0.62

foreign

-0.61

POSITIVE LOGITS

ĵĺ

0.77

 stairs

0.74

ISSION

0.71

Topic

0.69

ikuman

0.69

 seams

0.67

atche

0.66

ĪĴ

0.66

orthy

0.65

 sights

0.65

Activations Density 0.168%

phrases related to things being up-to-date or upcoming

phrases indicating trends or changes over time

No Comments

No Known Activations