INDEX

Explanations

locations or points in texts where a shift or transition occurs

oai_token-act-pair · gpt-3.5-turbo

phrases indicating location or circumstances

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

Hunter

-0.63

mage

-0.62

mob

-0.59

 trave

-0.58

Bug

-0.57

bug

-0.57

Abyss

-0.56

morph

-0.56

iencies

-0.56

 disapprove

-0.56

POSITIVE LOGITS

ESA

0.75

ulic

0.74

liest

0.72

ioned

0.71

anos

0.67

SPONSORED

0.66

ITAL

0.66

risis

0.65

essional

0.64

arters

0.64

Activations Density 0.054%

locations or points in texts where a shift or transition occurs

phrases indicating location or circumstances

No Comments

No Known Activations