INDEX

Explanations

references to user actions and interactions within a digital platform

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

amer

-0.72

 Baptist

-0.70

 Lutheran

-0.70

von

-0.63

stress

-0.61

hovah

-0.61

BJ

-0.61

 Tempest

-0.61

 Scand

-0.60

 Vaugh

-0.60

POSITIVE LOGITS

 interface

1.06

 interfaces

1.01

interface

0.96

pace

0.95

 Interface

0.92

base

0.91

cript

0.87

Agent

0.82

Interface

0.81

ifest

0.77

Activations Density 0.454%

references to user actions and interactions within a digital platform

No Comments

No Known Activations