INDEX

Explanations

specific instances of communication interactions, such as receiving calls, emails, letters, or notifications

oai_token-act-pair · gpt-3.5-turbo

terms related to communication and feedback mechanisms

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

occupied

-0.66

aple

-0.64

 Stru

-0.59

 Doors

-0.59

 Leilan

-0.57

 Wiki

-0.56

ndra

-0.55

 wells

-0.53

Vert

-0.53

 Tycoon

-0.53

POSITIVE LOGITS

 from

1.21

 FROM

1.06

from

1.03

 From

0.80

 congrat

0.77

 firsthand

0.75

From

0.75

 courtesy

0.72

reply

0.71

 compliments

0.71

Activations Density 0.333%

specific instances of communication interactions, such as receiving calls, emails, letters, or notifications

terms related to communication and feedback mechanisms

No Comments

No Known Activations