INDEX

Explanations

negative clauses indicating absence or denial

oai_token-act-pair · gpt-3.5-turbo

expressions indicating negation or denial

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

gnu

-0.68

Preview

-0.67

 Traps

-0.66

ULAR

-0.65

trap

-0.65

ships

-0.63

 Mushroom

-0.60

enos

-0.59

kind

-0.59

ãĥį

-0.58

POSITIVE LOGITS

wegian

1.11

nor

0.86

hovah

0.86

berto

0.78

 Lumpur

0.74

theless

0.73

 shalt

0.71

fleet

0.70

withstanding

0.67

onso

0.67

Activations Density 0.010%

negative clauses indicating absence or denial

expressions indicating negation or denial

No Comments

No Known Activations