INDEX

Explanations

specific adjectives or terms related to describing characteristics or behavior

oai_token-act-pair · gpt-3.5-turbo

phrases that include descriptors related to being "headed" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

jam

-0.78

vo

-0.70

 Blend

-0.66

ALS

-0.64

 Â®

-0.63

 Magazine

-0.58

 Remastered

-0.57

 recharge

-0.57

Mod

-0.56

 express

-0.56

POSITIVE LOGITS

headed

4.53

heading

2.24

head

2.12

heads

2.06

faced

1.71

footed

1.54

tailed

1.49

eyed

1.45

 headed

1.42

HEAD

1.38

Activations Density 0.014%

specific adjectives or terms related to describing characteristics or behavior

phrases that include descriptors related to being "headed" in various contexts

No Comments

No Known Activations