INDEX

Explanations

keywords or terms that stand out as important or central points

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

uum

-0.74

 Chatt

-0.72

 Lucia

-0.69

ews

-0.68

 Laksh

-0.67

Fif

-0.65

 Kitt

-0.63

Poe

-0.63

 Tsukuyomi

-0.62

ICAN

-0.62

POSITIVE LOGITS

stroke

1.33

stone

1.33

binding

1.19

*/(

1.18

stro

1.11

stones

1.05

bind

1.05

chain

1.04

pad

1.04

hole

1.02

Activations Density 0.553%

keywords or terms that stand out as important or central points

No Comments

No Known Activations