INDEX

Explanations

instances of the word "and" in various contexts

oai_token-act-pair · gpt-4o-mini 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

ĸļ

-0.74

advertising

-0.71

advertisement

-0.71

 procedure

-0.69

endum

-0.68

otion

-0.67

cade

-0.67

ridor

-0.67

store

-0.66

VERTISEMENT

-0.66

POSITIVE LOGITS

 Hispanics

1.08

 radicals

1.04

 Catholics

1.02

 independents

1.02

 Latinos

1.02

 communists

1.00

 anarchists

0.99

 minorities

0.98

Nazis

0.98

 nationalists

0.98

Activations Density 0.207%

instances of the word "and" in various contexts

No Comments

No Known Activations