INDEX

Explanations

instances where the word "will" is used to convey future actions or beliefs

oai_token-act-pair · gpt-3.5-turbo

expressions of future predictions or expectations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

eatured

-0.67

amped

-0.67

amps

-0.65

 Theme

-0.64

engineering

-0.63

sembly

-0.63

 Romeo

-0.62

Methods

-0.62

Reporting

-0.62

 Measures

-0.61

POSITIVE LOGITS

 gladly

1.12

 inevitably

1.09

be

1.08

 doubtless

1.08

 surely

1.07

 undoubtedly

1.07

 probably

1.05

 eventually

1.03

 likely

1.02

 never

0.97

Activations Density 0.215%

instances where the word "will" is used to convey future actions or beliefs

expressions of future predictions or expectations

No Comments

No Known Activations