INDEX

Explanations

words related to actions or events that happened or will happen

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

mare

-0.71

Post

-0.68

soever

-0.66

 vividly

-0.66

###

-0.61

,.

-0.61

Introduced

-0.61

Due

-0.61

pour

-0.60

Press

-0.60

POSITIVE LOGITS

bidden

1.00

geries

0.89

gotten

0.86

gery

0.81

ked

0.78

 example

0.78

 clarification

0.76

 awhile

0.76

ties

0.74

WARD

0.69

Activations Density 0.093%

words related to actions or events that happened or will happen

No Comments

No Known Activations