INDEX

Explanations

mention of political entities and policies

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

obin

-0.76

apult

-0.72

rolet

-0.68

]);

-0.68

Cause

-0.68

abo

-0.64

iously

-0.63

elta

-0.63

Detailed

-0.63

staking

-0.63

POSITIVE LOGITS

 however

1.08

 meanwhile

1.00

 moreover

0.95

there

0.90

 there

0.88

 where

0.83

where

0.80

 namely

0.70

 though

0.70

we

0.67

Activations Density 0.847%

mention of political entities and policies

No Comments

No Known Activations