INDEX

Explanations

phrases indicating addition, particularly in a negative context

oai_token-act-pair · gpt-3.5-turbo

phrases indicating additions or contributions to a situation or context

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

CHAT

-0.77

glers

-0.63

iker

-0.61

 Reporter

-0.60

cko

-0.59

rowth

-0.59

 Dane

-0.59

soever

-0.58

Released

-0.58

leground

-0.58

POSITIVE LOGITS

 compensate

0.73

otos

0.68

ventory

0.67

 enhance

0.67

 counteract

0.65

 scenery

0.65

 strengthen

0.64

 bolster

0.63

 prominence

0.63

 shore

0.63

Activations Density 0.063%

phrases indicating addition, particularly in a negative context

phrases indicating additions or contributions to a situation or context

No Comments

No Known Activations