INDEX

Explanations

proper nouns, specifically names of individuals or locations

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

âĶĢâĶĢ

-0.74

 ANGEL

-0.70

ktop

-0.70

BOX

-0.69

 VIDE

-0.66

 daytime

-0.66

 srfAttach

-0.62

 volume

-0.61

 Serie

-0.61

 PROG

-0.60

POSITIVE LOGITS

baugh

1.31

enberg

1.25

hoff

1.17

hart

1.14

ley

1.12

love

1.10

ingham

1.09

gren

1.08

berger

1.07

meyer

1.06

Activations Density 0.231%

proper nouns, specifically names of individuals or locations

No Comments

No Known Activations