INDEX

Explanations

political entities and conflicts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ILCS

-0.61

ank

-0.56

ullivan

-0.52

 Kara

-0.51

 pestic

-0.51

 Alley

-0.51

OL

-0.50

 hemor

-0.50

 charism

-0.50

 Kham

-0.50

POSITIVE LOGITS

 coincides

1.19

 outweigh

1.17

 coincided

1.15

 varies

1.13

 depends

1.10

 outwe

1.09

 exceeds

1.08

 reflects

1.06

 involves

1.06

 constitutes

1.04

Activations Density 3.729%

political entities and conflicts

No Comments

No Known Activations