INDEX

Explanations

mentions of food items or dishes

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

icus

-0.90

hips

-0.85

icals

-0.85

ica

-0.80

ical

-0.80

istically

-0.78

antha

-0.78

ists

-0.76

icum

-0.71

uments

-0.71

POSITIVE LOGITS

washer

1.15

fare

0.88

mares

0.85

robe

0.76

enger

0.74

conom

0.72

lete

0.71

onest

0.71

verages

0.71

ttes

0.70

Activations Density 0.180%

mentions of food items or dishes

No Comments

No Known Activations