INDEX

Explanations

locations or place names

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

gerald

-0.66

 adjourn

-0.64

REDACTED

-0.62

:{

-0.61

Ö¼

-0.59

glers

-0.58

 Craw

-0.57

OTOS

-0.55

EMENT

-0.55

 allowances

-0.54

POSITIVE LOGITS

ieri

0.82

ophon

0.80

ophone

0.79

encia

0.78

ocene

0.72

oise

0.67

dale

0.67

agne

0.67

 Nieto

0.67

quin

0.65

Activations Density 0.047%

locations or place names

No Comments

No Known Activations