INDEX

Explanations

medical-related terms or acronyms, specifically 'MD'

oai_token-act-pair · gpt-3.5-turbo

references to medical degrees or professionals, specifically "MD"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

acters

-0.90

stall

-0.86

=-=-=-=-=-=-=-=-

-0.79

asonable

-0.77

flies

-0.76

toggle

-0.76

nel

-0.74

ãĥ¼ãĥ³

-0.74

neau

-0.70

istical

-0.70

POSITIVE LOGITS

iamond

0.96

iate

0.84

aughter

0.80

MD

0.80

ouble

0.78

ORPG

0.77

IVER

0.74

MT

0.73

ription

0.73

insula

0.72

Activations Density 0.006%

medical-related terms or acronyms, specifically 'MD'

references to medical degrees or professionals, specifically "MD"

No Comments

No Known Activations