INDEX

Explanations

names of individuals

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

AAA

-0.71

LEASE

-0.62

ournal

-0.61

į

-0.60

ãĢĲ

-0.59

 Thumbnails

-0.58

ĻĤ

-0.57

ĺħ

-0.57

division

-0.55

give

-0.55

POSITIVE LOGITS

enegger

1.05

's

0.97

gren

0.93

quist

0.93

 testified

0.92

 himself

0.90

stein

0.89

berg

0.89

sey

0.88

hoff

0.88

Activations Density 3.474%

names of individuals

No Comments

No Known Activations