INDEX

Explanations

events or situations that gained significant attention or notoriety

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

tein

-1.02

 halves

-0.69

 recip

-0.65

 planner

-0.63

 Plan

-0.63

 Simulator

-0.60

ihad

-0.59

 woodland

-0.59

ampa

-0.58

 instructions

-0.57

POSITIVE LOGITS

ocl

0.90

rovers

0.90

 internationally

0.90

elight

0.87

worthy

0.86

Reviewer

0.84

isSpecialOrderable

0.83

 domestically

0.82

 endorsements

0.82

ously

0.80

Activations Density 1.441%

events or situations that gained significant attention or notoriety

No Comments

No Known Activations