INDEX

Explanations

phrases related to specific events and locations, possibly related to news or stories

oai_token-act-pair · gpt-3.5-turbo

introductory phrases and reference points in discussions

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

forth

-0.71

 superiors

-0.65

 acquies

-0.65

 thereof

-0.65

 opposite

-0.65

 equivalent

-0.65

 attest

-0.62

 hereafter

-0.60

nown

-0.60

 prec

-0.60

POSITIVE LOGITS

 âĢº

0.83

zbollah

0.80

 Vegan

0.74

ccording

0.72

 SHARES

0.71

 Bought

0.70

CLUS

0.70

inav

0.68

 Recap

0.68

 Expand

0.67

Activations Density 1.249%

phrases related to specific events and locations, possibly related to news or stories

introductory phrases and reference points in discussions

No Comments

No Known Activations