INDEX

Explanations

code snippets that involve looping through collections and performing summation calculations

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

 ancest

-0.69

ivities

-0.68

 Jinn

-0.68

andise

-0.68

actor

-0.67

ignore

-0.63

 parach

-0.63

itton

-0.61

ifact

-0.61

 props

-0.60

POSITIVE LOGITS

eenth

1.28

een

1.20

1.06

ecause

1.00

swick

0.92

0.89

0.85

nesday

0.84

0.80

pher

0.80

Activations Density 0.983%

code snippets that involve looping through collections and performing summation calculations

No Comments

No Known Activations