INDEX

Explanations

questions and discussions about specific topics

oai_token-act-pair · gpt-3.5-turbo

inquiries and expressions related to seeking or discussing information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 declass

-0.60

 interchange

-0.59

 blat

-0.57

 dwar

-0.57

 respectfully

-0.56

 tripled

-0.56

ascular

-0.54

 veter

-0.53

 glim

-0.53

 helicop

-0.53

POSITIVE LOGITS

 entail

0.68

Ca

0.64

ĨĴ

0.63

INGTON

0.63

allery

0.60

Mich

0.60

antic

0.60

supp

0.59

icz

0.57

Cry

0.56

Activations Density 0.091%

questions and discussions about specific topics

inquiries and expressions related to seeking or discussing information

No Comments

No Known Activations