INDEX

Explanations

names of politicians and legal terms

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

iday

-0.82

aries

-0.81

vier

-0.77

naire

-0.75

naires

-0.74

ering

-0.74

UTION

-0.73

venue

-0.71

engers

-0.69

olver

-0.67

POSITIVE LOGITS

teenth

1.22

teen

1.04

some

0.88

eva

0.76

isy

0.74

oshenko

0.71

ãĥ£

0.70

 externalToEVAOnly

0.69

Runner

0.69

 corners

0.69

Activations Density 2.638%

names of politicians and legal terms

No Comments

No Known Activations