INDEX

Explanations

names of people or entities followed by specific information or actions related to them

oai_token-act-pair · gpt-3.5-turbo

proper nouns and names

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

issance

-0.65

ascript

-0.59

taboola

-0.54

arlane

-0.53

oyer

-0.53

ruary

-0.53

ĨĴ

-0.52

 ACTIONS

-0.52

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.52

 arrang

-0.50

POSITIVE LOGITS

adesh

0.57

silver

0.55

utra

0.52

 Splash

0.52

ï¸ı

0.52

pload

0.50

road

0.49

urat

0.49

 swat

0.49

 helicopters

0.48

Activations Density 1.338%

names of people or entities followed by specific information or actions related to them

proper nouns and names

No Comments

No Known Activations