INDEX

Explanations

words related to conflict or tension

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

 Annotations

-0.60

 Converted

-0.58

 Aren

-0.56

 Dresden

-0.56

 incorrectly

-0.53

 Sind

-0.53

 Retail

-0.53

 Hendricks

-0.51

 Hots

-0.50

 pairs

-0.50

POSITIVE LOGITS

neck

0.90

iest

0.82

bowl

0.81

piece

0.78

worn

0.78

headed

0.75

bag

0.75

ament

0.73

wagon

0.73

tro

0.73

Activations Density 3.062%

words related to conflict or tension

No Comments

No Known Activations