INDEX

Explanations

personal experiences or stories

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

rules

-0.61

 electromagnetic

-0.60

 totality

-0.59

PTS

-0.59

 relevance

-0.59

 Impact

-0.58

 Plaint

-0.57

 Allied

-0.55

 Gale

-0.55

 Georgian

-0.55

POSITIVE LOGITS

'm

1.69

've

1.44

'll

1.23

 suppose

1.21

'd

1.21

 guess

1.18

am

1.15

ggy

1.11

nex

1.09

 dunno

1.08

Activations Density 0.903%

personal experiences or stories

No Comments

No Known Activations