INDEX

Explanations

specific words related to the concept of tuning, like “tuner” or “tuning”

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

IDA

-0.66

 Misty

-0.62

 Angels

-0.61

USE

-0.60

 online

-0.59

UNCH

-0.59

 probable

-0.57

 Danger

-0.57

 chloride

-0.57

heid

-0.57

POSITIVE LOGITS

nels

1.30

nel

1.27

eful

1.14

eness

1.10

nell

1.09

geon

1.06

eless

1.06

kered

0.99

edo

0.98

estamp

0.94

Activations Density 0.023%

specific words related to the concept of tuning, like “tuner” or “tuning”

No Comments

No Known Activations