INDEX

Explanations

references to people's names or pronouns in a conversational context

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

AAA

-0.76

Lex

-0.67

ãĥ¼ãĥ³

-0.63

fif

-0.58

lehem

-0.57

 Seym

-0.57

CAP

-0.56

SOC

-0.56

ELF

-0.55

NAT

-0.55

POSITIVE LOGITS

 himself

1.25

enegger

1.22

 testified

1.15

 admits

1.08

's

1.07

 joked

1.04

 concedes

1.03

 wrote

0.98

 says

0.97

 explained

0.96

Activations Density 2.048%

references to people's names or pronouns in a conversational context

No Comments

No Known Activations