INDEX

Explanations

specific acronyms or abbreviations

oai_token-act-pair · gpt-3.5-turbo

acronyms or abbreviations for various organizations and programs

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

roman

-0.84

angel

-0.79

gaard

-0.77

lying

-0.75

ord

-0.72

iary

-0.71

orio

-0.69

court

-0.69

hai

-0.67

ices

-0.67

POSITIVE LOGITS

ISM

1.04

LAB

1.04

FORM

1.04

ET

1.02

FOR

1.01

ELY

1.00

ER

0.99

IFIED

0.98

IS

0.98

IUM

0.97

Activations Density 0.202%

specific acronyms or abbreviations

acronyms or abbreviations for various organizations and programs

No Comments

No Known Activations