INDEX

Explanations

phrases related to quick summaries or conclusions

oai_token-act-pair · gpt-3.5-turbo

phrases relating to summaries or concluding statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

clad

-0.51

tnc

-0.51

 fame

-0.48

emale

-0.47

peg

-0.47

burning

-0.46

ãģ¾

-0.44

grown

-0.44

 respectively

-0.42

management

-0.42

POSITIVE LOGITS

etheless

0.67

nces

0.65

terness

0.59

 confir

0.58

 citiz

0.56

ccording

0.56

Ĥª

0.55

pse

0.55

 glim

0.53

 FANTASY

0.52

Activations Density 7.223%

phrases related to quick summaries or conclusions

phrases relating to summaries or concluding statements

No Comments

No Known Activations