INDEX

Explanations

personal pronouns

oai_token-act-pair · gpt-3.5-turbo

statements that refer to the speaker's personal experiences or opinions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

pires

-0.61

 externalToEVAOnly

-0.60

arching

-0.59

 arcs

-0.59

 Impact

-0.58

 vacancies

-0.58

irlf

-0.58

 Prelude

-0.57

GCC

-0.57

ategory

-0.57

POSITIVE LOGITS

 mean

1.31

 swear

1.26

 dunno

1.19

'm

1.13

 guess

1.08

 guarantee

1.07

 suppose

1.04

ain

1.00

ronic

0.99

 dare

0.99

Activations Density 0.206%

personal pronouns

statements that refer to the speaker's personal experiences or opinions

No Comments

No Known Activations