INDEX

Explanations

interesting or curious concepts, developments, or rulings mentioned 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hops

-0.93

ournaments

-0.83

hers

-0.81

asters

-0.81

ateurs

-0.79

ULTS

-0.77

casts

-0.77

assies

-0.75

agents

-0.74

casters

-0.74

POSITIVE LOGITS

 distinction

1.20

 combination

1.14

 tale

1.04

 complication

1.04

 caveat

1.03

 twist

1.01

 tactic

1.01

 paradox

1.01

 contradiction

1.00

 irony

1.00

Activations Density 0.157%

interesting or curious concepts, developments, or rulings mentioned in the text

No Comments

No Known Activations