INDEX

Explanations

the word "have" followed by a verb

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

osi

-0.65

 hostage

-0.61

 plaintiff

-0.59

ocol

-0.57

 muse

-0.56

 Creep

-0.53

 Cold

-0.52

 hawk

-0.52

 trope

-0.52

Mp

-0.52

POSITIVE LOGITS

 been

1.35

been

1.29

 gotten

1.25

gotten

1.09

 gone

1.07

 Been

1.03

 undergone

0.97

 taken

0.94

 listened

0.94

 begun

0.92

Activations Density 0.401%

the word "have" followed by a verb

No Comments

No Known Activations