INDEX

Explanations

phrases indicating physical injuries or conditions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

rir

-0.65

Reviewer

-0.63

amiya

-0.63

 characterized

-0.63

.–

-0.62

arthy

-0.61

hops

-0.61

ulative

-0.59

ghai

-0.58

POSITIVE LOGITS

 impunity

1.11

 dignity

1.02

draw

0.97

drawn

0.97

stood

0.88

rosis

0.85

 regard

0.85

 regards

0.85

holding

0.85

 respect

0.84

Activations Density 0.057%

phrases indicating physical injuries or conditions

No Comments

No Known Activations