INDEX

Explanations

internally related terms or phrases within a given context

oai_token-act-pair · gpt-3.5-turbo

references to internal and external operations or processes

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

 Mehran

-0.80

 Ortiz

-0.73

 Roses

-0.69

 Archangel

-0.68

 Cause

-0.65

 Royals

-0.64

 Gamble

-0.63

 Wallace

-0.62

Khe

-0.61

 Parenthood

-0.61

POSITIVE LOGITS

ationally

0.90

terness

0.87

vernment

0.87

 displaced

0.85

 combustion

0.84

umin

0.84

hari

0.84

tical

0.83

omorphic

0.83

ovo

0.83

Activations Density 0.018%

internally related terms or phrases within a given context

references to internal and external operations or processes

No Comments

No Known Activations