INDEX

Explanations

phrases related to commands or instructions

oai_token-act-pair · gpt-3.5-turbo

negation or conditional phrases in the text

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

 respectively

-0.71

 thereto

-0.62

 ..."

-0.60

 âĢ¦"

-0.59

.","

-0.57

 thereof

-0.56

Ïī

-0.53

 prest

-0.52

etc

-0.52

ÏĢ

-0.51

POSITIVE LOGITS

resa

1.06

odore

0.94

xiety

0.84

notations

0.69

romeda

0.67

swers

0.67

mosp

0.65

zbollah

0.62

withstanding

0.62

bidden

0.61

Activations Density 0.755%

phrases related to commands or instructions

negation or conditional phrases in the text

No Comments

No Known Activations