INDEX

Explanations

phrases emphasizing a specific point or explanation

oai_token-act-pair · gpt-3.5-turbo

the term "what" in various contexts, indicating a search for definitions, explanations, or justifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

robe

-0.83

ster

-0.61

 congrat

-0.60

rop

-0.60

 favour

-0.60

uttering

-0.59

ffe

-0.58

case

-0.57

trop

-0.56

gen

-0.55

POSITIVE LOGITS

soever

1.14

 happens

1.07

 happened

1.00

 happ

0.99

 sorts

0.90

 kinds

0.83

 transpired

0.81

 separates

0.80

 distinguishes

0.73

 constitutes

0.72

Activations Density 0.082%

phrases emphasizing a specific point or explanation

the term "what" in various contexts, indicating a search for definitions, explanations, or justifications

No Comments

No Known Activations