INDEX

Explanations

phrases related to personal experiences or opinions

oai_token-act-pair · gpt-3.5-turbo

statements indicating disappointment or unmet expectations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 periodically

-0.63

 subconscious

-0.62

onomy

-0.62

ynamic

-0.56

 Spectrum

-0.54

 undoubtedly

-0.54

agos

-0.54

 doubtless

-0.53

 gradually

-0.53

 echoed

-0.53

POSITIVE LOGITS

âĢ

1.51

 anymore

1.41

âľ

1.13

ãĢ

1.06

Â¨

1.05

âĸł

1.02

Ì¶

1.02

anything

1.01

 âĢ

1.00

 anything

1.00

Activations Density 0.622%

phrases related to personal experiences or opinions

statements indicating disappointment or unmet expectations

No Comments

No Known Activations