INDEX

Explanations

locations and directions

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

ISM

-0.81

BILL

-0.75

Organ

-0.74

Written

-0.72

gae

-0.71

alian

-0.70

UX

-0.70

lamm

-0.67

CD

-0.66

 arrang

-0.64

POSITIVE LOGITS

sighted

1.01

shore

0.97

 misses

0.92

entimes

0.77

 Bethlehem

0.77

 midnight

0.75

 proximity

0.73

ctic

0.72

 infancy

0.72

ctica

0.72

Activations Density 0.340%

locations and directions

No Comments

No Known Activations