INDEX

Explanations

the word "much"

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

İĭ

-0.84

etts

-0.78

emy

-0.75

ologies

-0.73

opers

-0.73

ades

-0.71

raid

-0.71

ATURES

-0.69

emies

-0.69

agate

-0.68

POSITIVE LOGITS

ado

1.02

 simpler

0.80

 else

0.79

NESS

0.75

body

0.74

 misinformation

0.70

 acclaim

0.70

 resemblance

0.69

 appreciated

0.69

 attention

0.69

Activations Density 0.059%

the word "much"

No Comments

No Known Activations