INDEX

Explanations

phrases related to contrasting statements or clauses

oai_token-act-pair · gpt-3.5-turbo

commas 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

è¦ļéĨĴ

-0.62

sth

-0.62

ode

-0.61

eers

-0.60

cial

-0.58

ttes

-0.58

nexus

-0.57

://

-0.57

ruction

-0.56

ze

-0.54

POSITIVE LOGITS

 unlike

1.19

 alas

1.16

 contrary

1.12

 despite

1.09

 although

1.05

hey

1.04

despite

0.99

 barring

0.97

 unsurprisingly

0.96

 unfortunately

0.95

Activations Density 0.085%

phrases related to contrasting statements or clauses

commas in the text

No Comments

No Known Activations