INDEX

Explanations

concepts related to comparing or discussing differences, often across various contexts

oai_token-act-pair · gpt-3.5-turbo

recurring references to geographical or demographic distributions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

nery

-0.78

IZ

-0.66

nas

-0.63

 Lauder

-0.63

etic

-0.63

 coincidence

-0.62

nel

-0.60

nels

-0.59

zo

-0.59

 partName

-0.57

POSITIVE LOGITS

ĨĴ

0.98

roads

0.95

urst

0.90

ĸļ

0.87

cedented

0.85

scl

0.82

eatures

0.80

artisan

0.77

 across

0.76

keye

0.74

Activations Density 0.024%

concepts related to comparing or discussing differences, often across various contexts

recurring references to geographical or demographic distributions

No Comments

No Known Activations