INDEX

Explanations

specific locations and technical terms related to system operations

oai_token-act-pair · gpt-3.5-turbo

references to specific sites or systems

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ĻĤ

-0.84

etheless

-0.81

axter

-0.81

äºĶ

-0.80

ounter

-0.78

iannopoulos

-0.74

resent

-0.73

issance

-0.71

 heel

-0.71

¥ŀ

-0.71

POSITIVE LOGITS

 Clause

1.06

 Line

1.06

 Selection

0.96

 Reports

0.93

 Runner

0.91

 Room

0.91

 Definition

0.88

 Lines

0.88

 Teams

0.88

 Facility

0.87

Activations Density 0.178%

specific locations and technical terms related to system operations

references to specific sites or systems

No Comments

No Known Activations