INDEX

Explanations

abbreviations or acronyms, particularly those containing "AB"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letters "AB" in close proximity

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

 Turtle

-0.74

 turtle

-0.66

velt

-0.66

 Sergei

-0.66

 Tycoon

-0.64

 stranded

-0.62

 prin

-0.61

neo

-0.61

 Marshal

-0.60

 Dmitry

-0.60

POSITIVE LOGITS

AB

4.13

AB

1.75

ABLE

1.74

IB

1.65

ab

1.57

SB

1.51

AF

1.51

AG

1.48

AZ

1.40

Ab

1.37

Activations Density 0.016%

abbreviations or acronyms, particularly those containing "AB"

occurrences of the letters "AB" in close proximity

No Comments

No Known Activations