INDEX

Explanations

phrases related to legal terms or governmental actions

oai_token-act-pair · gpt-3.5-turbo

terms related to restrictions or regulations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

cffffcc

-0.78

FORE

-0.67

>>>>

-0.67

 Samoa

-0.61

 Corsair

-0.60

 promise

-0.58

 salv

-0.58

 exha

-0.57

 heck

-0.56

Khe

-0.55

POSITIVE LOGITS

rict

1.15

ombat

1.00

cliffe

0.91

ional

0.89

ions

0.88

abulary

0.86

ory

0.84

rait

0.84

iveness

0.83

riad

0.83

Activations Density 0.018%

phrases related to legal terms or governmental actions

terms related to restrictions or regulations

No Comments

No Known Activations