INDEX

Explanations

Christmas-related terms

oai_token-act-pair · gpt-3.5-turbo

terms related to rules or regulations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enegger

-0.79

ITNESS

-0.79

POL

-0.71

arella

-0.68

 Healer

-0.66

chrom

-0.66

ryption

-0.65

MODE

-0.64

 GOODMAN

-0.63

 à¨

-0.63

POSITIVE LOGITS

ule

0.92

lette

0.87

kas

0.86

cules

0.85

tta

0.83

ttle

0.82

ffe

0.81

bum

0.81

pee

0.80

quet

0.80

Activations Density 0.016%

Christmas-related terms

terms related to rules or regulations

No Comments

No Known Activations