INDEX

Explanations

abbreviations, acronyms, and technical terms

oai_token-act-pair · gpt-3.5-turbo

occurrences of specific acronyms or technical notations within parentheses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 resur

-0.70

sav

-0.68

 hawk

-0.64

 affiliate

-0.61

 haunted

-0.61

 elector

-0.60

 marsh

-0.60

icy

-0.59

 electors

-0.59

 Belgian

-0.59

POSITIVE LOGITS

aka

1.14

BIP

1.11

FP

1.08

HUD

1.07

PP

1.07

SG

1.07

BAT

1.07

UGC

1.06

RNA

1.05

ICS

1.05

Activations Density 0.080%

abbreviations, acronyms, and technical terms

occurrences of specific acronyms or technical notations within parentheses

No Comments

No Known Activations