INDEX

Explanations

years and countries

oai_token-act-pair · gpt-3.5-turbo

numerical or identifier-related references in varied contexts

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

ancy

-0.72

acle

-0.63

acles

-0.62

lihood

-0.62

vious

-0.61

 predec

-0.60

versive

-0.60

ï¸ı

-0.59

atron

-0.58

 canopy

-0.58

POSITIVE LOGITS

oglu

0.72

 Emirates

0.69

merga

0.67

Ô

0.67

bia

0.66

 neighbours

0.65

isine

0.64

ÃŁ

0.63

 Senegal

0.63

 Belg

0.63

Activations Density 0.693%

years and countries

numerical or identifier-related references in varied contexts

No Comments

No Known Activations