INDEX

Explanations

adjectives related to the perceived strength or significance of actions or events

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

 attempting

-0.70

senal

-0.68

archment

-0.65

 careful

-0.64

Warning

-0.61

iling

-0.61

dding

-0.60

 arguing

-0.59

 submitting

-0.59

 Levant

-0.59

POSITIVE LOGITS

 happened

1.22

 happens

1.12

 hurts

1.11

 transpired

1.06

 seems

1.05

 occurred

1.03

 mattered

1.03

 feels

1.03

 boils

1.02

 took

0.99

Activations Density 0.093%

adjectives related to the perceived strength or significance of actions or events

No Comments

No Known Activations