INDEX

Explanations

dates in various formats

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 withd

-1.05

 challeng

-0.82

 grounding

-0.82

 inver

-0.82

 nesting

-0.82

 prototyp

-0.81

 cabbage

-0.75

 mosqu

-0.74

 compr

-0.74

 glim

-0.74

POSITIVE LOGITS

jpg

1.38

 Retrieved

1.20

gif

1.14

png

1.11

 Accessed

1.05

txt

1.03

exe

1.02

COM

1.02

zip

1.00

com

0.99

Activations Density 0.184%

dates in various formats

No Comments

No Known Activations