INDEX

Explanations

instances of the word "watching"

oai_token-act-pair · gpt-3.5-turbo

instances of the gerund "watching" and its past tense "watched."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ctrl

-0.84

interstitial

-0.81

 congr

-0.77

esan

-0.75

eno

-0.69

ascal

-0.69

inia

-0.68

ãĥ´

-0.67

enture

-0.67

phi

-0.66

POSITIVE LOGITS

 Watching

1.18

dogs

1.12

tower

1.07

dog

1.01

 watching

0.97

 watched

0.87

 helpless

0.87

 passively

0.81

opes

0.80

 closely

0.77

Activations Density 0.034%

instances of the word "watching"

instances of the gerund "watching" and its past tense "watched."

No Comments

No Known Activations