INDEX

Explanations

phrases related to a specific place or institute

oai_token-act-pair · gpt-3.5-turbo

variations of the word "talk" in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

============

-0.71

CES

-0.64

 Staples

-0.63

======

-0.62

================================================================

-0.62

isine

-0.61

 requ

-0.61

ãģŁ

-0.60

CA

-0.60

ACTED

-0.60

POSITIVE LOGITS

alk

1.37

owitz

1.04

enstein

1.04

reath

1.00

iewicz

0.97

edIn

0.93

enburg

0.91

erness

0.84

yip

0.83

ers

0.83

Activations Density 0.009%

phrases related to a specific place or institute

variations of the word "talk" in different contexts

No Comments

No Known Activations