INDEX

Explanations

descriptions of physical actions and dialogues

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

ancial

-0.76

Reviewer

-0.73

Sources

-0.72

conservancy

-0.72

NW

-0.71

national

-0.71

 stunts

-0.70

 nominees

-0.70

 Critics

-0.69

regate

-0.68

POSITIVE LOGITS

 grin

1.22

 nodded

1.19

 smir

1.17

 grinned

1.13

 smiled

1.12

 frown

1.12

 grinning

1.12

 murm

1.10

 nods

1.09

 smile

1.08

Activations Density 1.796%

descriptions of physical actions and dialogues

No Comments

No Known Activations