INDEX

Explanations

punctuation marks, especially question marks, in the text

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Opera

-0.77

 Marketable

-0.75

 raft

-0.69

 comprom

-0.67

theless

-0.67

 Tanzania

-0.67

 destro

-0.66

 Firefox

-0.66

 Tasman

-0.66

 Mirage

-0.66

POSITIVE LOGITS

ª

1.35

ł

1.14

Ĳ

1.11

Ĵ

1.09

¤

1.08

¹

1.06

ı

1.03

«

1.03

ĳ

1.03

Ļ

1.02

Activations Density 0.142%

punctuation marks, especially question marks, in the text

No Comments

No Known Activations