INDEX

Explanations

abbreviations or acronyms ending with "HS"

oai_token-act-pair · gpt-3.5-turbo

references to specific entities or organizations

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

fman

-0.81

taboola

-0.74

 Samoa

-0.69

osi

-0.67

dylib

-0.67

 awoken

-0.63

notations

-0.63

 Homo

-0.63

nesota

-0.63

tackle

-0.62

POSITIVE LOGITS

ocial

1.04

HS

1.00

IFT

0.94

INESS

0.91

ELF

0.91

TERN

0.89

YS

0.87

Ds

0.85

EMA

0.84

BC

0.83

Activations Density 0.015%

abbreviations or acronyms ending with "HS"

references to specific entities or organizations

No Comments

No Known Activations