INDEX

Explanations

mentions of government and its agencies

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

Rank

-0.76

 Junction

-0.74

sbm

-0.72

 Heaven

-0.71

uesday

-0.69

 Daylight

-0.69

 Velocity

-0.69

catentry

-0.68

uberty

-0.67

orkshire

-0.67

POSITIVE LOGITS

 agencies

1.04

 Accountability

0.99

 shutdown

0.98

 watchdog

0.95

 bureaucracy

0.94

 coffers

0.94

 agency

0.89

 contractor

0.85

 crackdown

0.82

 authorities

0.82

Activations Density 0.044%

mentions of government and its agencies

No Comments

No Known Activations