INDEX

Explanations

medical terminology and abbreviations

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

 Pell

-0.70

 takeoff

-0.67

 Peterson

-0.61

 Shepherd

-0.61

rules

-0.59

scl

-0.58

 cart

-0.58

 Fenrir

-0.57

 Jenner

-0.56

ont

-0.56

POSITIVE LOGITS

ptions

0.98

pta

0.96

cific

0.93

ccess

0.91

ificant

0.90

ruct

0.84

itutes

0.82

ption

0.81

ction

0.80

ctive

0.80

Activations Density 0.053%

medical terminology and abbreviations

No Comments

No Known Activations