INDEX

Explanations

mentions of wars

oai_token-act-pair · gpt-3.5-turbo

references to wars

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.2.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

AUT

-0.77

FORMATION

-0.70

OGR

-0.67

ostics

-0.65

Safety

-0.64

Asset

-0.64

UCT

-0.63

YL

-0.62

osis

-0.61

SOURCE

-0.61

POSITIVE LOGITS

hips

1.11

 wars

1.04

lords

1.00

bucks

0.96

hip

0.94

poons

0.91

riors

0.89

 waged

0.82

 raged

0.79

 battles

0.79

Activations Density 0.010%

mentions of wars

references to wars

No Comments

No Known Activations