INDEX

Explanations

phrases indicating addition or inclusion of information

oai_token-act-pair · gpt-3.5-turbo

imperative verbs suggesting action or instructions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

emale

-0.67

displayText

-0.67

PATH

-0.63

 wheelchair

-0.62

REF

-0.60

PRES

-0.60

mith

-0.60

 bottleneck

-0.59

SUP

-0.59

specified

-0.58

POSITIVE LOGITS

ings

0.90

 yourselves

0.77

anon

0.72

atio

0.72

antha

0.72

lihood

0.71

 Fasc

0.69

aring

0.69

 Quote

0.69

 Your

0.68

Activations Density 0.173%

phrases indicating addition or inclusion of information

imperative verbs suggesting action or instructions

No Comments

No Known Activations