INDEX

Explanations

mathematical equations and numerical values

oai_token-act-pair · gpt-3.5-turbo

equality or equivalence statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 livest

-0.90

 nodd

-0.84

eness

-0.82

imony

-0.79

wcs

-0.79

emouth

-0.76

itage

-0.74

etitive

-0.73

psey

-0.70

rov

-0.70

POSITIVE LOGITS

========

1.51

============

1.32

===

0.95

ãĥīãĥ©ãĤ´ãĥ³

0.82

nil

0.70

0.69

eric

0.69

ãĤ¨ãĥ«

0.68

[];

0.64

 Fasc

0.64

Activations Density 0.014%

mathematical equations and numerical values

equality or equivalence statements

No Comments

No Known Activations