INDEX

Explanations

mentions of locations and businesses, possibly related to a directory or contact information

oai_token-act-pair · gpt-3.5-turbo

numerical values and references to quantitative data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 analys

-1.14

foc

-1.04

 util

-0.98

 enqu

-0.97

 organis

-0.96

 whilst

-0.94

 analyse

-0.91

 scept

-0.91

 massively

-0.91

 emphas

-0.90

POSITIVE LOGITS

Copyright

1.74

Officials

1.66

Residents

1.64

Contact

1.61

RELATED

1.57

CLOSE

1.50

According

1.50

About

1.50

Mayor

1.49

Investigators

1.49

Activations Density 0.383%

mentions of locations and businesses, possibly related to a directory or contact information

numerical values and references to quantitative data

No Comments

No Known Activations