INDEX

Explanations

terms related to legal positions and titles

oai_token-act-pair · gpt-3.5-turbo

references to legal professionals and their titles

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

birth

-0.80

plant

-0.69

plates

-0.69

hook

-0.68

ften

-0.67

getting

-0.67

nings

-0.66

ãĥĩãĤ£

-0.65

Æ

-0.64

ãĤ·ãĥ£

-0.64

POSITIVE LOGITS

QC

1.10

arre

0.80

elling

0.76

UGH

0.75

RA

0.74

axter

0.72

WER

0.71

NB

0.71

rosse

0.70

ryce

0.69

Activations Density 0.014%

terms related to legal positions and titles

references to legal professionals and their titles

No Comments

No Known Activations