INDEX

Explanations

phrases indicating a premise or condition

oai_token-act-pair · gpt-3.5-turbo

the word "Given."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

CHA

-0.78

imming

-0.77

á

-0.74

ater

-0.72

Prev

-0.71

Ã¡

-0.71

ereo

-0.69

arser

-0.68

Ã¦

-0.68

utters

-0.68

POSITIVE LOGITS

stellar

0.81

McF

0.79

 precedence

0.69

 knowing

0.69

pend

0.69

sid

0.67

 undue

0.67

intendent

0.66

 scheduling

0.66

imaru

0.65

Activations Density 0.018%

phrases indicating a premise or condition

the word "Given."

No Comments

No Known Activations