INDEX

Explanations

references to triggering events or situations that may provoke strong reactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hemat

-0.78

apolis

-0.76

ensable

-0.74

apest

-0.73

ately

-0.72

nian

-0.71

esan

-0.69

cott

-0.68

 Cutler

-0.67

ijk

-0.67

POSITIVE LOGITS

 triggering

1.04

trigger

0.98

 warnings

0.95

 triggers

0.94

 trigger

0.87

 triggered

0.83

 warning

0.77

 Trigger

0.77

 puberty

0.74

 alerts

0.74

Activations Density 0.027%

references to triggering events or situations that may provoke strong reactions

No Comments

No Known Activations