INDEX

Explanations

mentions of resignations and removals from positions of authority

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

aceae

-0.81

natureconservancy

-0.78

 vantage

-0.71

ridges

-0.69

aphael

-0.69

ibaba

-0.67

iba

-0.66

antasy

-0.65

amins

-0.65

00200000

-0.64

POSITIVE LOGITS

 sacked

1.02

 abruptly

0.92

 disgr

0.87

 accusing

0.85

 whistlebl

0.85

 resign

0.84

 sergeant

0.82

 disgrace

0.82

 tenure

0.81

 overseeing

0.79

Activations Density 0.120%

mentions of resignations and removals from positions of authority

No Comments

No Known Activations