INDEX

Explanations

references to specific entities or topics within various contexts

oai_token-act-pair · gpt-3.5-turbo

keywords related to cultural and societal topics, particularly those that involve identity and significant events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 CLSID

-0.38

é¾įå

-0.36

Discussion

-0.35

ãĥł

-0.34

 Gemini

-0.34

 Thrust

-0.34

ãĥİ

-0.33

 Cunning

-0.33

later

-0.32

Lago

-0.31

POSITIVE LOGITS

terness

0.40

 arrang

0.38

natureconservancy

0.37

 âĢº

0.36

sych

0.35

atform

0.33

hes

0.33

bender

0.33

taboola

0.32

weed

0.32

Activations Density 3.042%

references to specific entities or topics within various contexts

keywords related to cultural and societal topics, particularly those that involve identity and significant events

No Comments

No Known Activations