INDEX

Explanations

initials of certain entities or brands followed by a numerical value, likely referring to specific terms or entities

oai_token-act-pair · gpt-3.5-turbo

abbreviations or acronyms related to specific entities or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

wich

-0.89

anism

-0.82

tics

-0.81

sburgh

-0.80

eous

-0.77

ertodd

-0.76

lies

-0.76

iform

-0.76

andowski

-0.73

Compat

-0.69

POSITIVE LOGITS

SS

0.93

BF

0.91

GI

0.90

GS

0.89

FS

0.87

PM

0.86

BB

0.86

BC

0.86

RR

0.85

WD

0.84

Activations Density 0.037%

initials of certain entities or brands followed by a numerical value, likely referring to specific terms or entities

abbreviations or acronyms related to specific entities or concepts

No Comments

No Known Activations