INDEX

Explanations

terms related to historical and political events

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

 ability

-0.81

 undet

-0.77

 coral

-0.75

 bloc

-0.74

 lifes

-0.74

 inclusion

-0.74

 carriage

-0.74

 hatch

-0.73

 payroll

-0.73

 covenant

-0.73

POSITIVE LOGITS

Advertisement

1.91

But

1.69

Instead

1.64

Perhaps

1.61

Yet

1.61

That

1.61

Fortunately

1.61

Then

1.61

When

1.60

Unfortunately

1.60

Activations Density 1.310%

terms related to historical and political events

No Comments

No Known Activations