INDEX

Explanations

words related to organizations or institutions

oai_token-act-pair · gpt-3.5-turbo

abbreviations or acronyms related to organizations, reports, and scientific entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ado

-0.65

gie

-0.62

 Gors

-0.61

 Elixir

-0.61

 Breitbart

-0.61

 puff

-0.60

Esp

-0.59

 Bohem

-0.58

Fn

-0.58

 compan

-0.58

POSITIVE LOGITS

KI

1.04

WD

0.96

ESS

0.94

FORM

0.93

KT

0.92

ET

0.90

ITE

0.90

RAG

0.89

amily

0.89

HE

0.89

Activations Density 0.186%

words related to organizations or institutions

abbreviations or acronyms related to organizations, reports, and scientific entities

No Comments

No Known Activations