INDEX

Explanations

words related to cutting or removing things

oai_token-act-pair · gpt-3.5-turbo

words related to actions of drinking or sipping

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

needs

-0.67

 Palest

-0.67

Translation

-0.66

 å¤

-0.65

 Nanto

-0.65

Zed

-0.63

DEV

-0.63

é¾

-0.62

 Siber

-0.61

scape

-0.61

POSITIVE LOGITS

ipped

0.94

olicy

0.89

ipping

0.88

psey

0.86

aukee

0.84

ips

0.84

avascript

0.83

otle

0.83

etition

0.83

untled

0.82

Activations Density 0.011%

words related to cutting or removing things

words related to actions of drinking or sipping

No Comments

No Known Activations