INDEX

Explanations

points of contention, controversy, or disputes within a given context or situation

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

 Doing

-0.76

 Writing

-0.69

YING

-0.68

izont

-0.68

ivating

-0.68

 Messages

-0.67

Maker

-0.65

onding

-0.64

 Saying

-0.63

 Aware

-0.63

POSITIVE LOGITS

 resembled

1.13

 resembles

1.12

 existed

1.11

 happened

1.09

 translates

1.05

 happens

1.04

 coincides

1.03

 belonged

1.02

 resided

0.98

 earns

0.97

Activations Density 0.166%

points of contention, controversy, or disputes within a given context or situation

No Comments

No Known Activations