INDEX

Explanations

words related to sorting or categorization

oai_token-act-pair · gpt-3.5-turbo

terms related to the organization and categorization of information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

zanne

-0.70

eele

-0.70

ewitness

-0.69

kee

-0.68

oral

-0.65

 abduct

-0.65

onde

-0.65

mosp

-0.64

iche

-0.64

 abduction

-0.63

POSITIVE LOGITS

 sorted

3.49

 sorting

2.45

Sort

1.71

 Sort

1.69

sort

1.64

Viz

1.37

 mull

1.10

 grouped

1.01

 filtering

0.99

ado

0.98

Activations Density 0.027%

words related to sorting or categorization

terms related to the organization and categorization of information

No Comments

No Known Activations