INDEX

Explanations

job titles, positions, or roles within organizations

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

 intervals

-0.70

urrencies

-0.69

 tongues

-0.69

å§«

-0.69

 instruments

-0.65

 hypotheses

-0.65

 pots

-0.64

 ACTIONS

-0.63

 rivers

-0.63

 Scouts

-0.63

POSITIVE LOGITS

 staffer

0.73

è£ħ

0.72

robe

0.69

stice

0.67

athon

0.67

book

0.67

ixtape

0.66

 similar

0.65

posium

0.65

 named

0.65

Activations Density 0.291%

job titles, positions, or roles within organizations

No Comments

No Known Activations