INDEX

Explanations

positive and negative descriptors

oai_token-act-pair · gpt-3.5-turbo

adjectives and phrases that describe mixed or conflicting evaluations of situations or policies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

udder

-0.76

 srfAttach

-0.67

ptives

-0.66

ľ

-0.64

REL

-0.63

 Alternatively

-0.63

 Lastly

-0.62

Alternatively

-0.62

 whichever

-0.62

 optionally

-0.62

POSITIVE LOGITS

 nature

1.56

ness

1.24

 situation

1.08

 similarities

1.08

 resemblance

1.05

 implications

1.02

 relationship

1.01

 attitude

1.01

liness

1.01

 predicament

0.99

Activations Density 0.509%

positive and negative descriptors

adjectives and phrases that describe mixed or conflicting evaluations of situations or policies

No Comments

No Known Activations