INDEX

Explanations

words related to illegal or unethical behavior

oai_token-act-pair · gpt-3.5-turbo

elements related to deception or dishonesty

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

hess

-0.48

 Reef

-0.47

liner

-0.44

 âĨĳ

-0.44

alties

-0.42

rouse

-0.42

ASA

-0.42

cheon

-0.42

ounces

-0.41

 Weston

-0.41

POSITIVE LOGITS

0.57

Â»

0.57

''

0.55

\)

0.55

ï¸ı

0.55

`,

0.52

,''

0.52

ACTED

0.47

[/

0.47

-)

0.46

Activations Density 1.041%

words related to illegal or unethical behavior

elements related to deception or dishonesty

No Comments

No Known Activations