INDEX

Explanations

title abbreviations or acronyms

oai_token-act-pair · gpt-3.5-turbo

the acronym "TA" and its associated contexts

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

stocks

-0.81

space

-0.80

lain

-0.79

meyer

-0.75

nen

-0.74

ously

-0.71

wagen

-0.70

espie

-0.70

papers

-0.69

nan

-0.68

POSITIVE LOGITS

KE

0.97

BILITY

0.91

VE

0.88

WA

0.87

WN

0.86

KER

0.85

ZI

0.84

pling

0.83

ZE

0.82

ILS

0.82

Activations Density 0.014%

title abbreviations or acronyms

the acronym "TA" and its associated contexts

No Comments

No Known Activations