INDEX

Explanations

references to medical professionals or doctors

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

 fracture

-0.74

perm

-0.67

 tense

-0.66

 prem

-0.65

pt

-0.64

 crate

-0.62

 sled

-0.62

 drive

-0.62

 consortium

-0.61

 flap

-0.61

POSITIVE LOGITS

 Joseph

1.06

 Kenneth

1.02

 William

1.01

 John

1.01

 Gerald

1.00

 Richard

0.98

 Samuel

0.98

essors

0.97

 Judith

0.96

 Timothy

0.95

Activations Density 1.218%

references to medical professionals or doctors

No Comments

No Known Activations