INDEX

Explanations

actors, movies, and other entities that have been nominated for awards

oai_token-act-pair · gpt-3.5-turbo

instances of the word "nominated" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

iegel

-0.70

seless

-0.70

 fiber

-0.65

pees

-0.62

ful

-0.61

 Flow

-0.61

rael

-0.58

gew

-0.58

sur

-0.58

ound

-0.57

POSITIVE LOGITS

 nominated

3.42

 nominate

3.06

 nominations

2.84

 nominees

2.62

 nominating

2.42

 nomination

2.27

 nominee

2.01

Nom

1.34

nom

1.33

 disqualified

1.23

Activations Density 0.019%

actors, movies, and other entities that have been nominated for awards

instances of the word "nominated" and its variations

No Comments

No Known Activations