INDEX

Explanations

references to historical timelines and relationships specifically related to certain staff positions

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

$$$$

-0.73

 knots

-0.70

erers

-0.70

 PLEASE

-0.69

FIG

-0.68

cellaneous

-0.67

eric

-0.66

facing

-0.65

isance

-0.64

ctive

-0.64

POSITIVE LOGITS

1.08

 previous

1.08

1.05

 inception

1.04

 childhood

1.03

Activations Density 3.128%

references to historical timelines and relationships specifically related to certain staff positions

No Comments

No Known Activations